Some Differences in Corporate Structure in Germany, Japan

Some Differences in Corporate Structure in

Germany, Japan, and the United States

Mark J. Roe

CONTENTS

INTRODUCTION .............................................. 1928

I. THE CLASSICAL ECONOMIC MODEL OF THE PUBLIC FIRM .............. 1933

II. GERMAN AND JAPANESE INFLUENCE THROUGH STOCKHOLDING:

SHARED CONTROL AND A FLAMTER AUTHORITY STRUCrURE ............ 1936

A. Large Firms and Large Voting Blocks ........................ 1936

B. Structured Interaction ................................... 1941

C. Size of Foreign Financial Intermediaries ...................... 1946

D. American Banking Regulation ............................. 1948

1. German Universal Banks in the United States? .............. 1949

2. Japanese Main Banks in the United States? ................ 1951

3. Main Banks and Universal Banks-The Potential for Control .... 1954

4. Main Banks and Universal Banks-The Suppression and

Capture of the Securities Market ........................ 1955

E. Other Regulatory Impediments ............................. 1956

1m. FINANCIAL EVOLUTION IN GERMANY AND JAPAN? . . . . . . . . . . . . . . . . . . . . 1958

A. The Evolutionary Argument ............................... 1958

B. Problems with the Evolutionary Argument ..................... 1959

1. Securitization ..................................... 1959

2. Autonomy ........................................ 1962

3. American Evolution ................................. 1965

t Professor, Columbia Law School. Theodor Baums, Bernard Black, Victor Brudney, Ronald Gilson,Victor Goldberg, Jeffrey Gordon, Robert Hamilton, Hideki Kanda, Klaus Hopt, Lou Lowenstein, SteveKaplan, Fritz Kilbler, Gerald Neuman, Hugh Patrick, J. Mark Ramseyer, Detlev Vagts, and Paul Windolfmade useful comments. Douglas Cardwell, Stacy Cooper, Marco Geromin, Burkard Gopfert, MegumikoIshida, Steven Lorenz, Martina Stoldt, Hans-Gert Vogel, and Jean-Claude Zerey assembled data, research,and translations of German and Japanese materials. The Bradley Foundation supported this research.

1927

1928 The Yale Law Journal [Vol. 102: 1927

4. Economic Evolution and the Cost of Capital ................ 1966

IV. POLITICAL EVOLUTION IN GERMANY AND JAPAN? . . . . . . . . . . . . . . . . . . . . . 1967

A. German "Populism". ................................... 1967B. Japanese Interest Groups ................................. 1972

V. SHARED AUTHORITY: WEAKNESSES AND STRENGTHS .................. 1977A. Problems with Management ............................... 1979

1. Increasing Accountability ............................. 1979

2. Monitoring by Multiple Intermediaries .................... 1980a. Multiple Blocks Checking Each Other ................. 1981

b. Multiple Blocks Impelling Action ..................... 1981c. Multiple Blocks and Power Sharing ................... 1982

3. Monitoring the Banks ................................ 1983

4. Personifying Shareholders ............................. 19845. Improving Decisions ................................. 1985

B. Problems with Financial Markets: Institutions as the Problem ....... 1987

1. Enhancing the Information Flow ........................ 1987

2. Improving Loan Markets .............................. 1987

C. Problems in Large-Scale Industry: Organization ................. 1988

VI. IMPLICATIONS FOR AMERICAN STRUCrURE ........................ 1989A. Germany and Japan as Blueprint? .......................... 1990

B. Path Dependence and the Futility of Legal Change? .............. 1992C. The Paradox of International Competition: Solution and Problem .... 1994

CONCLUSION ............................................... 1995

APPENDIX .................................................. 1998

INTRODUCTION

Ownership of the large public corporation in Germany and Japan differsmarkedly from that in the United States. Here, senior managers hold the reinsof power; scattered individuals and institutions own the stock. Americanfinancial intermediaries only recently have begun to project power intocorporate boardrooms; they have historically been dispersed, weak, anduninterested. In the German and Japanese large firm, in contrast, seniormanagers must share power with active intermediaries wielding the votes oflarge blocks of stock.

Corporate Structure

These differences and the different law, history, and politics that helpedproduce them cast doubt on current thinking about the origin of the largeAmerican public firm. True, firms had to tap vast pools of capital to reacheconomies of scale and, as shareholders diversified, scattered ownership shiftedpower to managers. While this is the dominant economic paradigm, it omitsa critical step: American shareholders could have diversified through powerfulintermediaries, but they did not. Corporate functions could have become furtherspecialized, with strong intermediaries sharing governance functions withmanagers, but they did not. Had such intermediaries developed in the UnitedStates, a flatter structure of shared power could have developed in the large

American firm, as it did in Germany and Japan. Elsewhere I have argued that

one must look beyond economics to explain the American results: America'spolitics of financial fragmentation, rooted in federalism, populism, and interestgroup pressures, pulverized American financial institutions, contributingheavily to the rise of the Berle-Means corporation.'

In Part I of this Article, I argue that even a brief comparison of corporateownership structures between Germany and Japan on the one hand, and the

United States on the other, poses problems for prevailing academic theories.The classical economic explanation for the public firm emphasizes scale

economies, shareholder diversification, and the divergence of managerial goalsfrom stockholder goals. The economic explanation would, if it were universal,tend to predict that nations with similar economies would have similarcorporate structures. There is a best way to make steel, and presumably there

is a best way to organize large steel firms. Thus, managerial incentivecompensation schemes, proxy fights, conglomerates, takeovers, and boards ofindependent outsiders, all of which reduce (or reflect) the costs of organizingthe large American public firm, should play a role in corporate governance in

all three nations. At a minimum, the absence of these features in the flatter,shared authority structure prevalent in Germany and Japan poses a challenge,because it shows that there is more than one way to deal with the large firm'sorganizational problems. And the differences in corporate structure suggest thatdiffering political histories, cultures, and paths of economic development mustbe added to the economic model to explain the differing structures. Theeconomic model must be down-sized and considered to be the special case thatarises if intermediaries are fragmented.

The classical economic model does not account for the German andJapanese corporate structure, whose basic features I present in Part II. Largefinancial intermediaries, usually banks, hold concentrated blocks of stock inmany foreign firms, including the very largest. Bankers and managers interact

1. Mark J. Roe, A Political Theory of American Corporate Finance, 91 COLUm. L. REv. 10 (1991)[hereinafter Roe, Political Theory]; Mark J. Roe, Political and Legal Restraints on Ownership and Control

of Public Companies, 27 J. FIN. ECON. 7 (1990).

192919931

The Yale Law Journal

in structured settings, where power has been shared between firms andfinanciers, unlike in the United States, where the chief executive officer hashistorically controlled the firm from the top of a hierarchy, without significantaccountability to financial intermediaries.

Comparison allows us to test the hypothesis that law and the politics thatunderlie it help determine corporate structure. If U.S. firms could readily adoptthe German or Japanese structure under U.S. law, we would have a partialcontradiction of the law and politics hypothesis. In fact, several laws bar U.S.banks from involvement in the governance of American firms. Most critically,the foreign banks have a national reach that would violate the McFadden Act;they are relatively much larger than U.S. banks, giving them the size neededto control large slices of large firms' capital structure. The Bank HoldingCompany Act of 1956, as usually interpreted, requires U.S. banks to bepassive, but the foreign banks are not passive. The Glass-Steagall Act limitsAmerican banks' securities business.

German national banks enter boardrooms by combining votes from stockthey own directly, stock in bank-controlled investment companies, andsecurities they vote for their brokerage customers. Combining banks, mutualfunds, ahd brokers in this way would violate the McFadden Act, the Glass-Steagall Act, and the Bank Holding Company Act. Japanese banks are alsoactive in corporate governance through truly national banks and, even thoughJapan has a Glass-Steagall Act (imposed by America after World War I) toseparate commercial banks from investment banks, Japanese regulatorseffectively repealed it by forcing savings and corporate financing into Japanesebanks.

These laws, necessary to fully understanding the American public firm,have been the San Andreas fault line in American corporate governance,historically severing America's largest financial institutions from its largestindustrial firms. While we'll never know whether American-style fragmentationwould have developed in the absence of legal roadblocks, American law wouldhave been sufficient to block both German- and Japanese-style financialinstitutions from developing here.

In Part III, I examine whether the differences in corporate structure areonly temporary because financial institutions in Germany and Japan are"behind" America and evolving without political pressure to resembleAmerican institutions. There are contrary trends abroad-some point towardless concentration and less institutional voice, some point to more voice or tostability. And there are weak trends toward greater concentration andinstitutional voice here in the United States. The mixed facts indicate financialevolution is uneven; we are evolving to look (a little) like them and they areevolving to look (a little) like us.

Even if foreign financial institutions clearly were fragmenting andAmerican institutions were concentrating, doubts about current corporate

1930 [Vol. 102: 1927

Corporate Structure

theories would persist. Foreign politics could have induced the foreign

evolution. In Part IV, we see that today's foreign structures depend partly on

the path they took from the foreign statist past, which militated toward

financial concentration. Today political forces not all that different from those

in America's past are weakening the intermediaries. In Germany, persistent

popular pressure has led to Parliamentary proposals to reduce the power of

banks over industrial stock. In Japan, efforts to transform credit-based financial

power, which is weakening, to stock-based boardroom power, have not been

fully successful yet, because the laws left by the post-World War II Americanoccupation preclude such a transformation, and interest group conflict-mostly

between bankers and brokers-tends to support the status quo. To transform

their financial system, the Japanese must dismantle the American framework.

They are beginning to do so, but they may fail.In Part V, I analyze the common elements in the German and Japanese

structural patterns and compare them to the U.S. structure in order to find

theoretical explanations for their survival despite their obvious weaknesses.

Institutional ownership discourages entrepreneurial leadership, risks conflicts

of interest, and may make adaption to change more difficult. Moreover, agency

problems may simply shift from the firm to the intermediary. Although a

simple inquiry might try to find the better corporate structure and anoint it the

preferred candidate, complications afflict such an inquiry. First, there isn't

much evidence that the foreign structures contribute greatly to better

performance. Second, a legal inquiry differs from an economic one. As long

as institutional engagement in corporate governance advantages some firms, we

have reason to regret some of our restrictions. Finding some value in the

foreign structures, however, does not mean we should require them here;

absent spillover effects, we should only allow firms that want them to adopt

them. Third, even if we came to regret our restrictions, we might not

necessarily demand their repeal or predict that their repeal would lead to rapid

restructuring, because the path-dependent development of American financial

institutions and corporate structures impedes change. Fourth, it might turn out

that neither the American fragmentation nor foreign concentration is optimal,

or it might turn out that any foreign successes are culturally dependent or

embedded in institutions (like labor relations or the regulatory system) that are

impossible to import. The comparative inquiry may just show that there are

more possibilities than the American one. Fifth, product market competition

in an internationalized economy is the best self-corrective. Whatever the source

of the recent success of foreign firms-and I suspect corporate structure is

tertiary-competition induces American firms to perform better.What are the possible foreign strengths? Although a corporate law

academic's first inquiry would likely be to see if the German and Japanese

corporate structure controls managerial agency costs well, the competitive

advantage of the foreign structure may lie not just in reducing what we think

193119931


of as agency costs, but in (1) changing the environment ofdecisiorlmaking-bringing more individuals and organizations to the tablewhen technologies and markets are changing too rapidly for a single CEO ora single firm to stay current; (2) improving the information flow to largestockholders in ways that a fragmented securities market cannot achieve; or (3)improving organizational performance by facilitating relation-specificinvestments without using a stultifying, large vertical organization and byproviding a matrix for decisionmaking across related organizations. Even iffuture research shows none of these hypotheses about Germany and Japan tobe true, or shows that the advantages of the true hypotheses are offset by theirdisadvantages, the study of the differences in corporate structure is worthwhilebecause it illustrates that the current American arrangements are not inevitable,but instead are highly dependent on the organization of financialintermediaries. We may never want to import foreign features, but once we seethat corporate structure is variable, we may be more willing to reexamine ourown system.

In Part VI, we see that in reexamining our system, we should not allowGermany and Japan to become blueprints for American financial institutions.The organization of banks is too important to be driven by debatable corporategovernance concerns. Nor is it clear that legal change would induce structuralchange. The gains, if any, from change may be small and might fail to exceedthe transaction costs of any change. Complex institutions are shaped by theirown history, including their regulatory history; legal permission may not leadinstitutions to take the newly-opened opportunities, at least not immediately.Moreover, CEO's are embedded in several markets and organizational featuresbeyond corporate governance relations with institutional shareholders. Theseother markets and organizational features may be strong enough to render anygains from corporate change small.

Finally, I conclude this review of corporate structures with a simple claim:How a nation organizes its financial institutions deeply determines its largefirm corporate ownership structure. In America, a deep gap separates financeand industry; in Germany and Japan such a fault line does not yet fullyseparate intermediaries from managers. Neither ownership fragmentation dueto scale economies nor the economic model can explain enough of thestructural differences. Corporate ownership structure is highly sensitive to theorganization of financial intermediaries, which in turn is highly sensitive topopular attitudes toward financial concentration and political fights.

1932 [Vol. 102: 1927

Corporate Structure

I. THE CLASSICAL ECONOMIC MODEL OF THE PUBLIC FIRM

Since Adolf Berle and Gardiner Means published The Modem Corporation

and Private Property2 in 1932, American corporate law academics' central

task has been to understand the separation of ownership from control in the

large American corporation. In the standard story, the rise of large-scale

production technologies at the end of the nineteenth century demanded huge

inputs of capital that could only be raised from far-flung investors who,

demanding diversification and liquidity, were only able to buy small pieces of

a firm's equity. Small holders distant from the enterprise lacked the

information, skill, and incentive to monitor managers, who became relatively

free from shareholder oversight.

This unwieldy structure yielded agency costs, described in the modem

reformulation as the sum of (1) bonding costs, as managers tried to assure that

they would do a good job; (2) monitoring costs, as shareholders tried to

oversee managers; and (3) a residual loss that could not be eliminated.3

Corporate law scholarship sought to understand these costs, to explain features

of the firm as either aligning managers' interests with those of shareholders or

as the residuum of unavoidable loss, and to recommend ways to further reduce

these costs. In equilibrium, there would be no further costs to eliminate. Any

observed costs would be either the bonding or monitoring costs, whose

elimination would cost more in increased residual loss, or the residual loss,

whose elimination would cost more in offsetting bonding or monitoring costs.Corporate academics sought to perfect the independent director, who is not

beholden to management. Other means to minimize agency costs have included

managerial labor markets, incentive compensation tied to stock price, capitalmarkets that denied bad managers access to capital, and high debt that

heightened fear of bankruptcy. Conglomerates in their ideal form centralized

information and checked managers;4 in their pathological form, they built

empires. In contractarian corporate thought, corporate law was a contract to

minimize agency costs. State corporate law provided a standard form for most

firms and allowed the remaining ones to tailor the contract to their specific

situations.5 Finally, some financial theorists explained shareholderpowerlessness as efficient; stockholders should not influence firm decisionsbecause their specialty was bearing risk, not decisionmaking.6

2. ADOLF BERLE & GARDINER MEANS, THE MODERN CORPORATION AND PRIVATE PROPERTY (1932).

3. The 1970's state-of-the-art was Michael C. Jensen & William H. Meckling, Theory of the Firm:

Managerial Behavior Agency Costs and Ownership Structure, 3 J. FIN. ECON. 305, 308 (1976).4. OLIVER E. WILLIAMSON, MARKETS AND HIERARCHIES: ANALYSIS AND ANTITRUST IMPLICATIONS

155-75 (1975).5. FRANK H. EASTERBROOK & DANIEL R. FISCHEL, THE ECONOMIC STRUCTURE OF CORPORATE LAW

34-35 (1991).6. Eugene F. Fama, Agency Problems and the Theory of the Firm, 88 J. POL. ECON. 288 (1980).

19331993]


By the 1980's, however, agency theorists were unsure whether Americahad achieved a long-run equilibrium in corporate structure in which the furtherreduction of agency costs had become unimportant. Many scholars argued thattakeovers cured large, residual agency costs because managers who deviatedtoo much from shareholders' interests risked a hostile takeover. Skepticsviewed some takeovers as resulting from the agency costs of errant managersexpanding their empires, but others argued that even these costs would beeliminated in another wave of takeovers that would break up the empires.7Regardless of whether takeovers cured or exacerbated agency problems, manycorporate law scholars viewed agency cost explanations as central tounderstanding the American corporation.8

Our classical economic model would imply that large firms in everyeconomy would face similar problems. The organization of large-scale industryin the United States generated agency problems and solutions; large-scaleindustry elsewhere should tend to develop the same organization, problems,and solutions. As foreign industry "caught up" with American industry,economies of scale would require huge inputs of capital from far-flunginvestors who would let power shift to managers, leading to a similarfragmented ownership, afflicted with similar managerial agency costs. Thealternatives to this prediction might be too horrible for corporate law scholarsto imagine. Corporate structure might not matter. Or, different governancesystems might yield different advantages and disadvantages, making it difficultfor one system to dominate the other. Or, each system might be efficient in itsown national context, indicating that there are several equally efficient waysto organize large-scale industry. Or, the U.S. governance system may havearisen and survived because U.S. laws and politics suppressed the obviousalternatives to the Berle-Means corporation.

The differences in corporate structure seem to contradict the classicaleconomic model, at least in its most simple form. Elsewhere I have advancedan additional hypothesis to explain the development of the large public firmin America.9 There is more than one way to move savings from householdsto industry. A securities market that produces fragmented ownership is onlyone way, and it is the prevailing way in America. But, funds could also move

7. Bernard S. Black, Bidder Overpayment in Takeovers, 41 STAN. L. REV. 597, 627-28 (1989).8. For overviews of the agency theory literature, see Michael C. Jensen & Clifford W. Smith,

Stockholder Manager and Creditor Interests: Applications of Agency Theory, in RECENT ADVANCES INCORPORATE FINANCE 93 (Edward I. Altman & Marti G. Subrahmanyam eds., 1985) and Clifford W. Smith,Agency Costs, in THE NEW PALGRAVE: FINANCE (John Eatwell et al. eds., 1989).

9. Roe, Political Theory, supra note 1. For related discussions, see WILLIAM G. OUCHI, THE M-FORMSOCIETY: How AMERICA CAN RECAPTURE THE COMPETITIVE EDGE 82-90 (1984); LESTER C. THUROW,THE ZERO SUM SOLUTION 164-65 (1985); Bernard S. Black, Shareholder Passivity Reexamined, 89 MICH.L. REV. 520 (1990); Ronald J. Gilson & Reinier H. Kraakman, Reinventing the Outside Director: AnAgenda for Institutional Investors, 43 STAN. L. REV. 863 (1991); Joseph A. Grundfest, Subordination ofAmerican Capital, 27 J. FIN. ECON. 89 (1990); Michael C. Jensen, Eclipse of the Public Corporation,HARV. BUs. REV., Sept.-Oct. 1989, at 61.

1934 [Vol. 102: 1927

Corporate Structure

from households to firms through large, powerful financial intermediaries that

control big blocks of stock in industry while providing their investors with

diversification and liquidity. Such powerful institutional intermediaries are

more important abroad than in America. Quite a bit of evidence indicates that

American politics deliberately fragmented financial institutions, their portfolios,

and their ability to aggregate stock into influential voting blocks. Thus, we

may hypothesize that if American political history had taken a different path,

financial intermediaries might have played a different role in the American

economy, and the structure of authority in the large public firm might also be

different.We cannot go back to 1890 and re-run American history without the

politics of financial fragmentation to see whether American financial

intermediaries and firms develop differently. We can, however, search abroad

for different foreign governance structures and see whether the differences

correlate with differences in financial regulation. In fact, the history of foreign

financial regulation differs from the American experience, supporting the

hypothesis that the politics of financial intermediation is a key determinant of

corporate structure. The economic model cannot alone explain foreign

structures and their differences with the American structures; it needs a

political theory of American corporate finance to provide an adequate

explanation.The contractarian version of the classical economic says corporate law is

a standard contract that shareholders and managers would adopt if they could

cheaply bargain over their relationship with the firm; or, in its normative form,

the contractarian version asserts that corporate law should conform to the

contract that parties would tend to negotiate. By focusing on state corporate

law, particularly Delaware corporate law and the ordering among shareholders,

boards of directors, and managers, we American corporate law scholars have

overlooked key laws that limit financial intermediaries' ability to control large

blocks of stock and to become involved in corporate governance. To co-opt the

political critique, proponents of the contractarian model must concede that

important mandatory rules are found, perhaps not in state corporate law, but

in the rules dictating the organization of financial intermediaries and their

relationship with large firms. They can then energize the normative form of the

economic model and criticize mandatory rules if the rules bar useful contracts.

The foreign differences thus cast doubt on the standard American

evolutionary paradigm-that corporate ownership mediated through securities

markets is the highest form of financial development, successfully providing

ownership, diversification, and liquidity in just the right proportions. If we

believed that financial fragmentation is economically (as opposed to politically)

inevitable then the economic model could survive as the sole explanation for

corporate structures. But the persistence thus far of differences in foreign

ownership structures suggests not only that more that one evolutionary path

19351993]


exists, but also that the recent increase in American institutional investors'corporate governance activity is probably the delayed result of our historicalsuppression of strong intermediaries.

The point here is not to displace the part of the economic model thatexplains the role of technology, economies of scale, and far-flung investors.Rather, the goal is to displace the next step in the model's reasoning-thatthese features must give rise to fragmented ownership, with distantshareholders and powerful managers. The classical economic model can stand,but only if down-sized to a local equilibrium arising when financial institutionsare fragmented. If a nation does not fragment its financial institutions, or if anation designs other means of corporate governance-by privileging labor, ormanagers, or one financial institution or another, for example-then differentorganizational structures may emerge. How a nation regulates capital'sdeployment will affect how firms are organized. The history of Americanfinancial fragmentation, the foreign systems, and the modest contemporaryconcentration of American institutional ownership all indicate not only thatthere is more than one way to move capital into large firms, but also that theAmerican-style public corporation is more a local custom than the result of aninevitable economic evolution.

II. GERMAN AND JAPANESE INFLUENCE THROUGH STOCKHOLDING:SHARED CONTROL AND A FLATrER AUTHORITY STRUCTURE

Differences in the ownership of large firms in Germany and Japan, on theone hand, and those in the United States, on the other, are startling. Stock inlarge German and Japanese firms is held in big voting blocks; frequently ahandful of institutional shareholders votes 20% of a firm's stock. In America,in contrast, even after the concentration and institutionalization of recent years,the largest five shareholders rarely together control as much as 5% of a largefirm's stock. Despite significant legal differences between Germany and Japan,both nations have ownership structures that provide for a sharing of power sothat no person or institution has complete control. The U.S. structure ofcorporate ownership focuses power in management, especially in the chiefexecutive officer.

A. Large Firms and Large Voting Blocks

In both Germany and Japan, but not in America, the typical large firm'sstock is held in concentrated voting blocks. CEO's of the ten largest Germanfirms usually face two or three big institutional blocks, each controlling 10%of the vote; CEO's of the ten largest American firms face no such biginstitutional blocks. CEO's at many large German companies face a smallgroup of institutional voting blocks that controls nearly half of the stock voted;

1936 [Vol. 102: 1927

1993] Corporate Structure 1937

the equivalent figure in the United States is only 5%. Table I presents the

voting blocks for typical large German firms; Table II presents GM's

ownership structure, which is typical for the large American firm,'" as well

as the voting structures for the largest automakers in Germany and Japan.

TABLE I. Aggregate Voting Blocks of LargestThree German Banks (1986)"

Rank and Name Percent of Percent of Percent of Percent of Total Percent

of Company Shares Shares Shares Shares of Shares

Voted at Held by Held by Held by Held by

Meeting Deutsche Dresdner Commerz Three Largest

Bank Bank bank Banks

1. Siemens 60.64 17.64 10.74 4.14 32.52

2. Daimler-Benz 81.02 41.80 18.78 1.07 61.66

3. Volkswagen 50.13 2.94 3.70 1.33 7.98

4. Bayer 53.18 30.82 16.91 6.77 54.50

5. BASF 55.40 28.07 17.43 6.18 51.68

6. Hoechst 57.73 14.79 16.92 31.60 63.48

9. VEBA 50.24 19.99 23.08 5.85 48.92

10. Table XIII shows similar structures for other large American firms. For a complete breakdown of

institutional ownership for the largest 10 American firms, see Carolyn Kay Brancato et al., InstitutionalInvestor Concentration of Economic Power: A Study of Institutional Holdings and Voting Authority in U.S.Publicly Held Corporations app. 2 (Sept. 12, 1991) (unpublished study, on file with author).

11. Source: Amo Gottschalk, Der Stimmrechtseinfluss der Banken in den Aktiondirsversammlungen der

Grossunternehmen, 5 WSI-MrIrEILUNGEN 294, 298 (1988). Although Gottschalk's data-from 1986-is

not current, it is the most recent data available. This table combines the German banks' votes from

brokerage stock, directly-owned stock, and, apparently, mutual fund stock.


TABLE I. Top Five Institutional Voting Blocks in GM, Daimler-Benz, andToyota1

2

Daimler-Benz Toyota General Motors

Institutional Percent of Institutional Percent of Institutional Percent ofManager Shares Manager Shares Manager Shares

Deutsche Bank 41.80 Sakura Bank 4.9 Mich. St. Treas. 1.42

Dresdner Bank 18.78 Sanwa Bank 4.9 Bernstein Stanford 1.28

Commerzbank 12.24 Tokai Bank 4.9 Wells Fargo 1.20

Sonst. Kredit 4.41 Nippon Life 3.8 CREF 0.96

Bayerische L-Bank 1.16 Long-Term Cred. Bk. 3.1 Bankers Trust NY 0.88

Top 5 Shareholders 78.39 21.6 5.74

Top 25 Shareholders N.A. N.A. 13.93

German bankers' voting power comes from direct ownership of stock,from control over investment companies, and, most importantly, from authorityto vote stock that the bank's brokerage customers own but deposit with thebank. Although the German banks do not own a large amount of stock directlyfor their own benefit, when they deploy their capital in stock ownership, theydeploy it in big blocks. As Table VIII in the Appendix shows, in the onehundred largest German industrial enterprises, banks directly own for their ownaccounts twenty-two blocks of at least 5% of the outstanding stock. Banks alsovote the small one or two percent blocks owned by bank-sponsored investmentcompanies.' 3 Finally, banks vote the brokerage stock they hold as custodian.Typically, individual investors deposit the stock they own with their bank, and,unless the owner gives the bank special instructions, the bank votes thecustodial shares. German banks often hold as custodian and vote more than10% of the stock of a large company. 4

How does the ability to vote large blocks of stock, some of which the bankowns for itself and most of which it holds as custodian, allow the bank to

12. Source: CDA Investment Technologies, 13(0 Institutional Portfolios, Spectrum InstitutionalPortfolios (database for year-end 1990 GM data, assembled by Riverside Economic Associates); Gottschalk,supra 11, at 298 (data for Daimler-Benz' voting blocks at 1986 shareholders' meeting); JAPAN COMPANYHANDBOOK 790 (T6y6 Keizai Inc., 1992) (data for Toyota's voting blocks). The GM and Toyota data aretypical for the very largest American and Japanese firms. Daimler-Benz' largest vote holder is atypicallylarge, due to Deutsche Banks' atypically large block of stock that it owns for the bank's own account. TheDaimler-Benz data is based on votes actually cast at the 1986 shareholders' meeting; GM and Toyota dataare based on stock owned with full voting authority, irrespective of whether the holder voted the stock. TheGM data is from 1990, Daimler-Benz from 1986, and Toyota from 1992.

13. Gottschalk, supra note I1, at 295.14. Compare Table I and Gottschalk, supra note 11, with Table VIII; see also Detlev F. Vagts,

Reforming the "Modem" Corporation: Perspectives from the German, 80 HARV. L. REv. 23,53-58 (1966);Hermann H. Kallfass, The American Corporation and the Institutional Investor: Are There Lessons FromAbroad?, 3 COLUM. Bus. L. REV. 775, 782-83 (1988).

Corporate Structure

influence the firm? In the large German firm, stockholders elect half of thesupervisory board, employees appoint the other half' 5 The supervisory boardappoints a management board and approves major corporate decisions; themanagement board handles day-to-day decisions. German bankers use theirvotes to elect bank nominees to the supervisory boards of ninety-six of the onehundred largest German firms, and in fourteen cases, a banker chairs thesupervisory board.' 6 Although no single bank generally controls an industrialfirm (Deutsche Bank's control of Daimler-Benz is exceptional), together thethree German large banks can, if they act in unison, dominate the shareholdingside of the supervisory board.' 7

To an American observer, this control over large voting blocks, rather thancontrol over credit, is the biggest difference between the German and theAmerican structure. 8 Although bank loans helped create close relationshipsbetween bankers and managers, the importance of the "house bank" debtor-creditor relationship has faded,' 9 and German firms no longer rely on a singlebank for credit. German banks control the proxy system, not monopolisticsources of credit.

The ownership of large firms in Japan is roughly analogous to that inGermany. Large Japanese firms typically belong to a keiretsu, a group of firmsand intermediaries that own some of each others' stock, usually aggregatingto ownership of half of each other's stock.20 A main bank owns 5% of thestock of the keiretsu's industrial firms, which in turn own some stock in themain bank. Generally, four other banks and insurers own blocks of stock in theindustrial firms, roughly equal to 5% of the outstanding shares, thus creatinga latent five-holder coalition with 20% of the outstanding stock. Japanese largefirms show a persistent pattern of concentrated ownership for the past twenty-five years, as Table III shows, a pattern of concentration that prevails, not justfor small or mid-sized firms, but for Japan's very largest-Toyota, Nissan,Matsushita, and Mitsubishi. In America, in contrast, the five largest holders in

15. Some employee members represent white collar workers, some represent blue collar workers, andsome are union representatives.

16. DEUTSCHER BUNDESTAG, ACHTEs HAUPTGUTACHTEN DER MONOPOLKOMMISSION, DRUCKSACHE

11/7582, 11. Wahlperiode 202-06 (July 16, 1990) (investigative report by Monopoly Commission ofGerman Parliament); see also THEODOR BAUMS, BANKS AND CORPORATE CONTROL 29-30 (Berkeley Lawand Economics Working Paper No. 91-1, 1991).

17. Although Deutsche Bank is the largest bank, Commerzbank and Dresdner together have roughlythe same number of votes.

18. True, German bankers do sit on some boards where they lack votes, as do American bankers. Thecontrast with the American system is not here, but in the many firms where a German banker votes 15%or more of the stock of a firm on whose board a bank employee sits. If American institutions controlledsuch big blocks of stock, they would also sit on the board and be taken seriously by senior managers.

19. See BAUMS, supra note 16, at 9.20. Michael L. Gerlach, Keiretsu Organization in the Japanese Economy, in POLITICS AND

PRODUCTIVITY: THE REAL STORY OF WHY JAPAN WORKS 141, 159 (Chalmers Johnson et al. eds., 1989).

19391993]

The Yale Law Journal [Vol. 102: 1927

It~c m- 0 m C> m0 r- 0'

-- - - -l - -~ - - t-

6 z - i i 6 6 \ 6 -:

q Ci

N C14 Cl- - - - - C

It n 1f 0 m c

Cl (Cl ClqC ClR ClCn l l

C) Cl Cl C Cl Cl Cl Cl ClR

\R It0'' Ox C '0 OR 00

0' CCi= C') '0 Ci C' C')C' Ci

no- C-- 00 (n ~I)

IC' It C') -2 08 0. C2 02 C')C

> (' C l 'i C t - ~ C4--Nx C 4 0- Cl4 N -

t-- Cl C-- Cl t-- 000 0 -\ Cl4 0'0 - oo0000 0000 (3\ C,

1940

00

(N.\Z

LU =

<CO

CIS

e> 0

>1

co

0,-x

cd

1 0

o

0

0

z

'0

z .z

r<.0


the aggregate rarely own as much as 5% of the outstanding stock of the largestfirms, the amount the single largest owner typically has in Japan.2

The 20% owned by five shareholders in the largest Japanese firms isbigger than any five-shareholder block in the twenty-five largest Americanfirms.2" In GM, Exxon, and IBM, the largest twenty-five shareholders voteless stock than the largest five stockholders in Japan's largest firms.' Ifgroups with twenty-five members could coordinate as well as groups with five,this would be a difference of little importance, but they cannot.24

Thus the first difference in corporate ownership structure is theconcentrated voting blocks in the largest German and Japanese firms, and theirabsence in American firms.

B. Structured Interaction

German and Japanese CEO's regularly interact with the large owners. InGerman firms, the supervisory board appoints a managerial board to a five-year term and loosely reviews both firm and management performance,typically two to four times a year. Although the CEO and the rest of themanagerial board handle day-to-day decisions, they must report to thesupervisory board, which the CEO may not join, much less dominate.'

Not only do German managers not control the proxy machinery, but it isdoubtful that they can even lawfully make a proxy solicitation,26 a solicitation

21. For examples, see Tables II and XIII. See generally Brancato et al., supra note 10, at app. 2.22. If we examine smaller American firms, however, we see levels of concentration in stock ownership

that begin to approach the concentration levels typically found in Germany and Japan, although Americanlevels still fail to exceed the concentration in the very largest Japanese firms. Compare Tables III and XIwith Paul Clyde, The Institutional Shareholder as a Monitor of Management tbl. II (June 1991)(unpublished manuscript, on file with author) (showing that top five financial institutions own 18.4% ofstock in sample of 511 American companies); see also Harold Demsetz & Kenneth Lehn, The Structureof Corporate Ownership: Causes and Consequences, 93 J. POL. ECON. 1155, 1165-66 (1985) (showing thattop 20 stockholders control, on average, 37.7% of outstanding stock in sample of 511 Americancompanies).

23. Tables II and XIII show an American 5-shareholder ownership level of only 5% and a 25-institutional investor ownership level of 13.93% (GM), 11.47% (Exxon), 13.54% (IBM), and 12.89% (GE),an ownership level which is typical of the largest American industrial firms. See generally Brancato et. al,supra note 10, at app. 2.

24. Moreover, such groups of 25 face legal problems in coordinating. Black, supra note 9, at 536(noting that proxy rules hinder coordination among investors); Roe, Political Theory, supra note 1(discussing legal barriers to acquisitions of large blocks of stock by financial institutions). Some obstacles,however, were recently reduced. Regulation of Communications Among Shareholders, Exchange ActRelease No. 34-31,326 [Current] Fed. Sec. L. Rep. (CCH) 85,051 (Oct. 15, 1992); 57 Fed. Reg. 48,276(1992) (to be codified at 17 C.F.R. §§ 240, 249) (reducing obstacles to exchange of views amongshareholders).

25. 2 ERNST GESSLER Er AL., AKTIENGESETZ 138-39 (1974) (commentary on § 105 of Germansecurities law). The managerial board's ("Vorstand's") leader, or "speaker," is first among equals. Severallarge firms are private firms (GmbH for Gesellschaft mit beschriinkter Haftung, or "firm with limitedliability"), whose stock is not traded. These firms tend to have founding families who seek to maintaininfluence.

26. Alfred F. Conard, The Supervision of Corporate Management: A Comparison of Developments inEuropean Community and United States Law, 82 MICH. L. REv. 1459, 1470 (1984).


process American managers dominate. Instead, German managers must filterproxy solicitations through bankers, who vote their own stock, their mutualfunds' stock, and their customers' custodial stock. The banks obtain fifteen-month revocable proxies from their brokerage customers, and prior to votingthe custodial stock, inform their customers of their intentions, giving theircustomers an opportunity to instruct them to vote differently.' Rationalapathy should lead most shareholders to ignore the solicitations, and, in fact,customers rarely disapprove of their bank's recommendation. 28 Bankerscontrol the proxy machinery and hence elect the stockholding side of manysupervisory boards.

It would, however, be easy to exaggerate the power of the Germansupervisory board. First, translation of the German board's name, Aufsichtsrat,as "supervisory board," while linguistically correct, does not quite reflect theboard's authority, which is less than that of a supervisor, a position one wouldordinarily suspect would yield authority to supervise. A better translation mightbe "advisory board," with "advisory" meaning more than just gratuitousconsultation but a power similar to that of the U.S. Senate to advise andconsent to treaties and appointments, which yields consultation and influencebut not supervisory control.

For example, the supervisory board cannot remove the managerial boardat will during its five-year term. Moreover, the managerial board co-opts someof the supervisory board's formal authority when a vacancy arises on thesupervisory board, in a manner similar to, but weaker than, the AmericanCEO's ability to name the board, by suggesting to the supervisory board whoshould fill the vacancy.29 German codetermination also induces shareholderrepresentatives to want the supervisory board to supervise less than itotherwise would, because a powerful supervisory board would enhance theauthority of labor's representatives on the board. The bankers generally preferto take their chances with the managerial board. True, if there were aboardroom confrontation, the banker-shareholders could defeat the employeesby relying on two features of German corporate law: First, the chair of thesupervisory board is always a member of the shareholding side of the boardand can cast the deciding vote in a tie; and, second, shareholders can, with asuper-majority vote, send directions to the managerial board, bypassing thesupervisory board. 30 Thus, the supervisory board provides shareholders with

27. AKTIENGESEIZ (AKTG), §§ 135(l)-(2) (1991) (German corporations code) (authorizing revocablegeneral proxy to banks).

28. Sprachlose Eigentilmer: Aktiondre nehmen viel zu selten ihre Rechte als Inhaber von Unternehmenwahr, DIE ZEIT, June 7, 1991 (stating that small shareholders cede virtually all voting rights to large banks);see also Gottschalk, supra note 11, at 296-97; Wir sind Gerngesehene Berater, DER SPIEGEL, Sept. 4, 1989,at 45, 48.

29. Paul Windolf, Codetermination and the Marketfor Corporate Control in the European Community,22 ECON. & SOC'Y 137, 140-41 (1993).

30. Id. at 143.

1942 [Vol. 102: 1927

Corporate Structure

influence, although not control, in corporate governance. Managers still havethe upper hand, but the tilt is not nearly as pro-managerial as it has historicallybeen in the United States.

Although the German pattern follows what an American corporate lawscholar would predict-that large stockholders would have board seats andinfluence-the Japanese pattern does not. Despite concentrated voting blocks,the boards of Japanese firms are typically made up of insiders, formally electedby stockholders but usually appointed by the CEO, with at most the occasionaladvice and consent of the large stockholders. Stockholders usually do notappear on the board absent a crisis. Thus, the pattern that an Americancorporate law scholar would predict-that large stockholders would have boardseats and influence-occurs only in crisis, or through an informal mechanism.

An informal mechanism for stockholder influence does exist in Japan. Thefinancial intermediaries' leaders interact with the managers of firms in monthlykeiretsu Presidents' Council meetings, which resemble a second board,analogous to the German supervisory board. Even though votes are not takenat the meetings and participants do not direct one another, individual presidentsfeel constrained by the consensus opinions of the council, largely because thecouncil members are capable people who collectively control much of thestock in the president's firm. Council members are said to be consulted onmajor decisions, as when a CEO chooses a successor.31

Our knowledge of the details of the monthly Presidents' Council meetingsis vague. Minutes are not leaked for us to examine and agendas for councilaction are not printed in the Japanese press. Some reports indicate that businessis discussed32 while others say the meetings are purely social,33 paving theway to do business later. In the following discussion, I take the Presidents'Council to include both the meetings and the information and decisionmakingchannels created through social interactions.

No single council member controls enough stock to control the others(although historically bank domination of credit yielded the bank control). Yet,since a group of these stockholders could exert control, no one member canwithstand the ire of a coalition of the others; moreover, in a culture that valuesconsensus, no one member should be willing to risk the ire of the others:

31. See v. CARL KESTER, JAPANESE TAKEOVERS: THE GLOBAL CONTEST FOR CORPORATE CONTROL

69 (1991) (describing role of keiretsu as including discussion of top personnel appointments); Charles A.Anderson, Corporate Directors in Japan, HARV. Bus. REV., May-June 1984, at 30, 32 (stating that ratherthan choosing a new CEO, incumbent CEO consults with the institutional owners of large blocks of firm'sstock and debt); cf. James B. Treece, 9 to 5 Japanese Style, Bus. WK., Dec. 28, 1992, at 20 (discussingsubtle, indirect communication through grapevine in Japanese firms).

32. MASAHIKO AOKI, THE ECONOMIC ANALYSIS OF THE JAPANESE FIRM 3, 12 (1984) (describing

Presidents' Council meetings as involving business discussions).33. Michael L. Gerlach, Twilight of the Keiretsu?-A Critical Assessment, 18 J. JAPANESE STUD. 79,

80-81 (1992).

19931 1943


These [council] meetings are not organs of decisionmaking in thesense that a majority vote would carry the day, but they aremanifestations of the very dynamic process of consensus. Views areexchanged, opinions heard, and actions reciprocally adjusted, more onan ad hoe basis than in terms of binding policy ....

The Presidents' Council is not a hierarchical command structure, but a forumfor communication and perhaps a collegial monitoring of near-equals.

The power Japanese banks historically exercised over credit complicatesany inquiry into the influence of institutional stockholders. When Japanesefirms were rapidly expanding after the Second World War, they sought newfunds from the banks, through which Japan channeled credit. The banks' powerto cut firms off from funding for new projects yielded the banks sufficientinfluence, irrespective of their stockholding.

Both stock and debt are relevant in the prevailing models, which seedelegated monitoring among Japanese banks, by which six banks buy stock in,and make loans to, several firms. The banker-group assigns each borrowingfirm a "main bank" to which the other banks delegate monitoring tasks. Inthese models, the main bank speaks with the authority of both creditors andstockholders owning 20% of the firm's stock."

The joint causation problem does not make stock irrelevant and, if banks'power over credit allocation continues to weaken, we shall see whether stockalone gives them influence. There is reason to believe that, although individualstockholders are powerless, intermediaries have influence through their stock.Although Japanese culture may deter large stockholders from formally firingthe CEO, mild criticism in a Presidents' Council meeting may shape actionsin a culture that values harmony. Bankers do send in directors when seriousproblems arise and performance weakens.36 The CEO may stay in place, butwith diluted authority, or the CEO may be "promoted" to the less powerfulposition of chairperson. In 1992, the CEO's of 14% of all Japan's publiccompanies left, about one-third of them involuntarily.37 Moreover, when

34. ROBERT J. BALLON & IWAO TOMITA, THE FINANCIAL BEHAVIOR OF JAPANESE CORPORATIONS68 (1988); see also MICHAEL L. GERLACH, ALLIANCE CAPITALISM: THE SOCIAL ORGANIZATION OFJAPANESE BUSINESS 108 (1992) (describing power of Presidents' Council members as implicit influencethat is continually negotiated); Anderson, supra note 31, at 30 (stating that Japanese corporate governance"appears to take place behind the scenes between the senior corporate official and the major institutionalshareowners").

35. Paul Sheard, The Main Bank System and Corporate Monitoring and Control in Japan, 11 J. ECON.BEHAV. & ORGANIZATION 399 (1989).

36. Steven N. Kaplan & Bernadette Alcamo Minton, "Outside" Intervention in Japanese Companies:Its Determinants and Its Implications for Managers (Sept. 1992) (unpublished manuscript, on file withauthor); Randall Morck & Masao Nakamura, Banks and Corporate Control in Japan (Sept. 1, 1992)(unpublished manuscript, on file with author).

37. ...And in Japan, FORTUNE, Feb. 22, 1993, at 10 (reporting ousters, although not explicitly tyingousters to stockholder initiatives). Although Japanese cross-ownership could insulate presidents from ouster,see John C. Coffee, Jr., Liquidity Versus Control: The Institutional Investor as Corporate Monitor, 91COLUM. L. REV. 1277, 1298-99 & n.82 (1991), the tumover and ousters indicate the insulation is

1944 [Vol. 102: 1927

Corporate Structure

informality fails and confrontation ensues, stockholding banks have takencontrol of the board and dismissed the CEO. When a scandal in a largedepartment store chain implicated the CEO, who uncharacteristically refusedto resign, the Mitsui Presidents' Council decided to replace him, and thebanker on the store's board engineered the coup de grace.38 Formal action isprobably usually unnecessary; in a small group, large shareholders should haveonly to raise their voice to assert influence.39

Other incidents, such as Japanese greenmail-twenty-two instances in the1980's-and a successful hostile takeover, indirectly suggest shareholderpower.40 Moreover, Japanese managers face fractious annual stockholdermeetings, which they control reasonably well because they go to the meetingsarmed with proxies from their large shareholders.4' Obviously, if the largestockholders denied managers the proxies, managers would face more difficultannual meetings. Finally, stockholding institutions unhappy with managers canthreaten to sell their stock, leaving managers at the risk of a takeover. Not onlyhave Japanese commentators described this potential as an important theoreticalsource of stock-based influence,42 but financial intermediaries have publiclymade such threats.43

In both the German and the Japanese large firm, structured interactionbetween stockholding intermediaries and corporate managers enables bankersto influence managers' actions, but not to control them completely.' InGermany, managers can ally with employees, who hold half of the seats on theboard. Since the large voting blocks are typically split among a handful ofbanks, the intermediaries must form a coalition to challenge an alliancebetween managers and employees. In Japan, banks are not represented on theboard, except in crisis. And, since bank-controlled stock is typically splitamong a handful of banks, a main bank must form a coalition to act. The

incomplete. If cross-holdings were fully protective, they would not allow roughly one out of every 20CEO's to be involuntarily ousted in a single year. The insulation hypothesis may confuse intent with effect;cross-holdings arose primarily because managers wanted to stave off takeovers and uncertainty, but friendly5% shareholders turn unfriendly when results are poor. Or the insulation hypothesis might have seemedwarranted when the Japanese economy was humming along and good results shielded even second-ratemanagers from inquiry.

38. GERLACH, supra note 34, at 111-13.39. Japanese corporate law poses no barrier, because it allows shareholders to remove directors without

cause at any time. AOKI, supra note 32, at 10.40. KESTER, supra note 31, at 17, 247-48.41. Merton J. Peck, The Large Japanese Corporation, in THE U.S. BUSINESS CORPORATION: AN

INSTITUTION IN TRANSITION 21, 22 (John R. Meyer & James M. Gustafson eds., 1988).42. Kunio Ito, M&A to Kabushik Mochiai no Honshitsu, KINYU J., Dec. 1989, at 11.43. Insurance companies have, in recent years, been unhappy with dividend payouts and have

threatened to dump the stock of companies that fail to increase their dividends. The observed threats arethe tip of an iceberg-of uncertain size--of private influence. See Stephen D. Prowse, The Structure ofCorporate Ownership in Japan, 47 J. FIN. 1121, 1138-40 (1992). The insurers' demands may not arisedirectly from firms' misuse of cash, but from the insurers' need to receive larger dividends to comply withinsurance regulations. I offer this item to show shareholder power, not to show that it is always used toimprove corporate governance.

44. There are exceptions, such as Deutsche Bank's dominant 28% block in Daimler-Benz.

19451993]


resulting authority structure is flatter than the one in the United States becausethe CEO seems unable to dominate decisionmaking. And since no singleintermediary can dominate, but must form a coalition to amass a dominatingblock of stock, the foreign governance structure is not equivalent to replacingan American-style dominating CEO with a dominating institution. Thus,structured interaction, without a shift in day-to-day control, is the second majorstructural difference between the foreign firms and the American.

C. Size of Foreign Financial Intermediaries

The third major structural difference is in the size of the financialintermediaries. The largest German and Japanese financial institutions play amuch bigger role in their economies than the largest American intermediariesplay in the American economy. Size and strength facilitate holding big blocksof stock, a holding that is difficult to achieve here in relatively smallerintermediaries. 45

The largest three German banks have assets equal to 36% of German grossnational product. The assets of the largest three American banks, in contrast,are equal to only 7% of the American GNP, making the German banks aboutfive times "stronger" than the largest American banks. Similarly, the largestbanks in Japan, a country with a GNP about 60% of that of America, areseveral times larger than their American counterparts.

TABLE IV. Summary of Bank Size in Germany, Japan, and America46

Country Assets of Assets of Foreign American Relative Relative Ratio of

Largest Largest GNP (3) GNP (4) Size of Size of Foreign Size

Three Three Foreign American to American

Foreign American Banks (5) Banks (6) Size

Banks (1) Banks (2) [(1)1(3)] [(2)/(4)] [(5)/(6)]

Germany $600B $424B $1.65T $5.5T 0.36 0.08 4.50

Japan $1,345B $424B $3.44T $5.5T 0.39 0.08 4.88

45. The largest three American banks are much weaker players in the American economy than thelargest German and Japanese banks are in their respective economies. See Table IV. These disparities instrength persist even if we compare each nation's largest ten banks. See Table V. Although currentexchange rates disfavor the dollar, using bank assets as a percentage of GNP, as in Table IV, eliminatesthe exchange rate distortion.

46. Source: Calculated from Table V.

1946 [Vol. 102: 1927


TABLE V. Assets of Ten Largest German, Japanese, and American Banks(in Billions of Dollars)47

Ten Largest German Banks Ten Largest Japanese Banks Ten Largest American Banks

Deutsche Bank 267 Dai-Ichi Kangyo 435 Citicorp 216

Dresdner Bank 189 Sumitomo 407 Bank America 110

Commerzbank 144 Fuji 403 Chase Manhattan 98

Bayerische V-Bank 138 Mitsubishi 392 J.P. Morgan 93

D. Genossen-Bank 137 Sanwa 387 Security Pacific 85

Westdeutsche 137 Industrial Bank of Japan 285 Chemical Bank 73

Bayerische H&W 117 Tokai 246 NCNB 65

Bayerische L-Bank 114 Bank of Tokyo 228 Bankers Trust NY 64

Nord-d Landesbank 83 Mitsubishi Trust 226 Manufacturers Hanover 62

Sudwest-d L-Bank 59 Norinchukin 224 Wells Fargo 56

Total 1,385 Total 3,233 Total 922

47. Source: Adapted from The Global Service 500: The 100 Largest Commercial Banking Companies,FORTUNE, Aug. 26, 1991, at 174-75.

Because American banks are relatively small, inserting the largest mutual fund complexes and thelargest insurers into Table V would make a difference, but not much of one:

TABLE Va. Assets of Ten Largest American Banks, Insurers, and Mutual Fund Complexes (1990, 1992)

Financial Institution Assets

Citicorp $217B

Fidelity $164B

Prudential $133BBank America $110B

Merrill Lynch $107B

MetLife $103B

Chase Manhattan $98B

J.P. Morgan $93B

Vanguard $93B

Security Pacific $85B

Total $1,202B

Source of asset size of insurers and mutual funds: CAROLYN KAY BRANCATO, INSTITUTIONAL INVESTORS:A WIDELY DIVERSE PRESENCE IN CORPORATE GOVERNANCE, tbls. 9, 12 (Background Paper Prepared forColumbia Institutional Investor Project, Center for Law and Economics, Feb. 25, 1993). The mutual fundnumbers are somewhat overstated because they are from 1992, while all other numbers are from 1990, andmutual funds grew from 1990 to 1992. Adding American pension funds would not make a big difference.TIAA-CREF, the largest pension fund with $99 billion worth of assets, would make it to the bottom of thelist. Calpers, the next largest pension fund, would not. Top 200 Pension Funds/Sponsors, PENSIONS &INVESTMENTS, Jan. 20, 1992, at 20. Expanding Table V to include insurers would add Allianz to theGerman list ($119 billion), but would not change the Japanese list. The Global Service 500, FORTUNE, Aug.24, 1992, at 215.


The relatively small size of American financial institutions, however, makesit harder for them to buy a large block of stock in GM, Exxon, or IBM thanit is for the foreign banks to buy 28% of Daimler-Benz or 5% of Toyota.

Let us summarize. Concentrated blocks, shared authority, and bigintermediaries are the three key structural differences.

Although we do not know exactly how the foreign structures work, and thedata is too thin to indicate whether they work well, we must not miss the mainpoint here. The foreign corporate structures differ greatly from those prevailingin America. German and Japanese senior managers share power with largefinancial intermediaries, which own and vote big blocks of stock and are activein corporate governance, formally through supervisory boards in Germany, andinformally through Presidents' Councils in Japan. The survival of these foreignfirms over several decades suggests that the classical economic model of thefirm must be reinterpreted as a special case in the U.S. setting because firmscan prosper with different governance structures. These differences appear tocorrelate, not so much with differences in economic task, but with differencesin the organization of financial intermediaries. Concentrated blocks, sharedauthority, and powerful intermediaries are not only uncommon here inAmerica; they have also been, as we next see, illegal.

D. American Banking Regulation

The organization of financial intermediaries deeply determines the structureof corporate ownership, and law deeply determines the organization offinancial intermediaries. We can see this syllogism vividly illustrated by testingwhether or not U.S. firms and intermediaries could imitate the foreignstructures without violating America's basic laws.

Size is central. Large American banks play a role in the Americaneconomy equal to only one-quarter of the role played by large banks inGermany and Japan.48 This difference in size correlates with law. Americanlegal restrictions have historically kept American banks small and weak, bybanning them from operating nationally, entering commerce, affiliating withinvestment banks, equity mutual funds, or insurers, or from coordinatingstockholdings with these other intermediaries.

The National Bank Act of 1863 historically confined national banks to asingle location, and the McFadden Act of 1927 only allowed branches ofnational banks to the extent state law permitted.49 Although states may permitout-of-state banks to open local branches, until recently they did not, and even

48. Table IV.49. National Bank Act of 1863, ch. 106, § 44, 13 Stat. 99 (1863) (codified as amended at 12 U.S.C.

§ 38 (1988)); McFadden Act, ch. 191, § 1, 44 Stat. 1224 (1927) (codified as amended at 12 U.S.C. § 36(1988)).

1948 [Vol. 102: 1927


the recent permissions are incomplete. Although federal law could overridestate law and permit interstate branching, it does not. America still lacks atruly national banking system like that of other nations, including Germany andJapan.

American banks have also faced product limits. The Glass-Steagall Act hashistorically denied banks a securities business and close affiliation withinvestment banks and, until recently, mutual funds.50 The Bank HoldingCompany Act prohibits affiliation with insurers and fine-tunes Glass-Steagallby prohibiting bank affiliation with nonbanks, except passive ownership of nomore than 5% of a nonbank's stock.

Lastly, American deposit insurance historically encouraged weak bankcapitalization, which limited banks' ability to make large equity investments.Bank managers can raise the private value of bank stock by keeping theirequity thin and displacing some of the risks of bank failure onto the public.51

By encouraging low levels of capital, extensive deposit insurance has mademany banks too weak to own much stock. If deposit insurance were lessextensive, banks would be pressed to raise more equity, either by market forcesor by new bank regulation, to attract uninsured deposits and to avoid bankruns.

52

1. German Universal Banks in the United States?

If U.S. banks tried to imitate German banks, they would smash into nearlyevery U.S. financial regulation. German banks not only have a national scopethat would violate McFadden's geographic restrictions 3 but they also holdnearly two dozen large positions in the one hundred largest German firms andthus would violate both the U.S. ban on bank stock ownership and the 5%

50. Act of June 16, 1933, ch. 89, 48 Stat. 162 (codified as amended in scattered sections of 12U.S.C.). Bank sponsorship of mutual funds, historically banned and only very recently accommodated underbanking law, is exploding, despite the fact that some of it is still in a legal gray area. Leslie Wayne,Questions on Bank Sales of Funds, N.Y. TIMES, Dec. 31, 1992, at Cl (noting explosive sales growth);Prospectus for Vista Capital Growth Fund 8 (Feb. 28, 1992) (on file with author) (prospectus of equitymutual fund sponsored by Chase Manhattan Bank concedes cloudy legality under Glass-Steagall of banksponsorship). Other restrictions are slowly eroding, particularly those that confine banks to the narrowbusiness of commercial banking. Nevertheless, the historical restrictions have important continuing effectsbecause they made American banks weak, and the weak banks will need years to evolve into stronger ones.

51. Thus, while the historical sequence was that Congress prohibited bank stock ownership first (in.1863 by the National Bank Act and in 1933 by Glass-Steagall) and then added deposit insurance (in 1934),we could think of the justification for the prohibition as coming in the reverse order. Once Congressestablished extensive deposit insurance, it also had to control the banks' level of risk and prohibiting stockownership was part of that control.

52. Recent regulatory initiatives forcing banks to increase their equity have been successful. Whetherunregulated banks would increase their own equity to enable them to hold stock in other firms woulddepend on whether the costs of holding big blocks of equity-such as lost liquidity and increasedrisk-outweighed the benefits. I outline some of the considerations below in Part V. There is, however, oneunnerving fact in favor of equity: stock persistently yields a higher retum than debt. Andrew Abel, TheEquity Premium Puzzle, FED. RES. BANK PHIL. MONTHLY REv., Sept. 1991, at 3. ,

53. McFadden Act, Pub. L. No. 69-639, 44 Stat. 1224, 1228 (1927).


limit on bank affiliate ownership of a nonbanking firm's stock.54 In addition,the largest German banks are also the largest brokerage houses, and theycontrol the proxy machinery, which would violate the historical productrestrictions of the Glass-Steagall Act. If U.S. banks were to act as brokers,stock exchange rules would prohibit them from voting their customers' stockon anything important,55 while German bankers can vote their customers'stock on anything at all, and they do.

German banks combine modest direct holdings of stock with extensiveholdings of custodial stock. American banks might try to imitate them bycombining modest stock holdings in affiliates with their extensive holdings oftrust stock.5 6 The American banks would face several obstacles, the first ofwhich arises from trust law, which tends to induce the hyper-fragmentation ofportfolios, to reward passivity, and to reinforce America's dominant investmentnorms of passivity.57

Trust law is a sufficient bar, but not the only one. A formal combinationof holding company stock and trust stock via the bank giving a proxy to theholding company would probably violate the Bank Holding Company Act,which prohibits bank affiliates from controlling more than 5% of a firm'sstock.58 An informal combination via consciously parallel voting would, byitself and if the banks did nothing more, seem to be within the limits of theAct. But its success would still depend on how aggressive the regulators wouldbe in applying interpretations that required holding companies to hold stockpassively. During an earlier manifestation of bank power through trust stock,Congress held extensive hearings and, through the vehicle of Wright Patman'sHouse Banking Committee, castigated banker power derived from owningstock in trust accounts and implicitly threatened banks with political costs ifthey used trust stock to exert influence in corporate governance. 59 Althoughwe can only speculate as to whether these political forces would arise againif banks tried to assert influence, we know that they did once arise and induced

54. See Table VIII.55. See, e.g., NEW YORK STOCK EXCHANGE, LISTED COMPANY MANUAL I402.06(D) (looseleaf 1990),

analyzed in Black, supra note 9, at 560-61; see also Conard, supra note 26, at 1469-70. Technically,American banks cannot be regulated as brokers, but they nevertheless were not permitted, until veryrecently, to operate brokerage firms, making the "freedom" from broker-dealer regulation moot. Again, theAmerican restrictions are partly historical; these features of Glass-Steagall are eroding.

56. The Glass-Steagall Act prohibits banks from themselves acquiring any stock.57. See Black, supra note 9, at 526-30; Jeffrey N. Gordon, The Puzzling Persistence of the

Constrained Prudent Man Rule, N.Y.U. L. REv. 52, 96-99 (1987); Mark J. Roe, Institutional Fiduciariesin the Boardroom, in INSTITUTIONAL INVESTING: CHALLENGES AND RESPONSIBILITIES IN THE 21ST

CENTURY 292 (Arnold W. Sametz & James L. Bicksler eds., 1991); Mark J. Roe, The Modem Corporationand Private Pensions (Apr. 1993) (unpublished manuscript, on file with author) [hereinafter Roe, ModemCorporation].

58. Bank Holding Company Act of 1956, ch. 240, §4(c)(6), 70 Stat. 133 (1956) (codified as amendedat 12 U.S.C. § 1843(c)(6) (1988)).

59. STAFF OF SUBCOMM. ON DOMESTIC FINANCE OF HOUSE COMM. ON BANKING AND CURRENCY,

89TH CONG., 2D SESS., REPORT ON BANK STOCK OWNERSHIP AND CONTROL 10 (Comm. Print 1966).

1950 (Vol. 102: 1927

Corporate Structure

prominent practitioners to warn banks not to use trust fund stock to buildcontrol blocks."

In any case, had American banks during most of this century tried toimitate the Germans, they would have failed since they have been barred fromoperating nationally, from entering the securities business, from using affiliatesto take blocks above 5%, and probably also from using trust stock to be active.McFadden, Glass-Steagall, the Bank Holding Company Act, and trust law eachseem sufficient to have deterred American bankers, who, if they tried to imitatethe German bankers, would have had to run their banks from jail.

2. Japanese Main Banks in the United States?

Japanese bank and insurer groups commonly control 20% of the stock ofa large firm, intervene during crises, and meet with managers during monthlyPresidents' Council meetings where, according to some,61 they reachconsensus on direction and operations. Could this system work under U.S. law,as some commentators assert, because no bank in the group owns more than5% of a firm's stock, the amount permitted a bank holding company-but nota bank-under U.S. banking law? In other words, can American bank holdingcompanies be active in corporate governance and take control in crisis as longas they limit their stock ownership to 5% of a firm's outstanding shares?62

American "main" banks would suffer from the geographic and productlimitations that make them puny compared to Japanese banks. As Table Vshows, while the assets of Japan's ten largest banks are over $3 trillion, theassets of America's ten largest banks do not total even $1 trillion-and theAmerican economy is larger than Japan's. No U.S. bank has the financialstrength needed to purchase a 5% stake in GM easily or to extend a huge loanwithout heavy syndication. The risk to a small American bank taking a bigslice of a large industrial firm's capital was historically heightened bygeographic restraints, which led to underdiversified assets and deposits. Abigger Japanese bank can purchase a 5% stake in Toyota without incurring thesame risks.63

Small size is enough to make law a complete barrier to an American mainbank system or America's largest firms. The Glass-Steagall Act adds to this

60. Raymond A. Enstam & Harry P. Kamen, Control and the Institutional Investor, 23 Bus. LAW. 289,290-91 & n.14 (1968); A. A. Sommer, Jr., Who's "In Control?"-SEC, 21 Bus. LAW. 559, 570 (1966).

61. See supra notes 32-34 and accompanying text.62. I've heard this argument at a number of conferences. Coffee in particular argues that because the

5% limit on stock ownership is common to both the U.S. and Japan, legal differences are unimportant.Coffee, supra note 37, at 1295.

63. See Table X. Nor would the ability of one bank to take a big block be likely to suffice. There isreason to think that multiple blocks make the system work, meaning that an American main bank systemmight need a half-dozen or dozen financial firms with the strength to take big blocks in multiple industrialfirms.

19931 1951


barrier by blocking the bank from owning stock anyway. And the BankHolding Company adds yet more to this barrier by encouraging bankerpassivity in wielding stock that affiliated companies own.64 The Act beginsby proscribing not just control of an industrial firm, but (subject to exceptions)ownership or control of any voting stock in an industrial firm.

Except as otherwise provided in this chapter, no bank holdingcompany shall-(1) after May 9, 1956, acquire direct or indirect ownership or controlof any voting share of any company which is not a bank .... 65

In the most aggressive interpretation of the prohibition, Citicorp would violatesection 4(a) by accepting an irrevocable proxy to vote a single share of GMstock. The Act then carves out an exception: the holding company may own"shares of any company which do not include more than 5 per centum of theoutstanding voting shares of such company...., 66 But, arguably, a holdingcompany that combined affiliate stock with trust stock by owning 5% andtaking proxies for another 15% of the outstanding shares would "acquire...indirect... control of... voting share[s] of any company which is not abank"' 67 in excess of the 5% exemption, and thereby would violate the Act.

An American banking group could try to overcome the 5% barrier byparallel action. Five banks and insurers could each own 5%; a main bankwould nominate directors; and all would vote their 5% for the nominees. Thisparallel action, even if coordinated with interbank discussion, would complywith the words of the statute but still face two banking law problems, onegeneral and one specific. (It would face securities law and other problems too).First, Congress castigated American banks as recently as the 1960's for havingmore subtle means of influence, leading lawyers to recommend that banks keepaway from such boardroom power.68 Second, the Federal Reserve Boardrejected a roughly similar proposal as banned by the Act:

[I]nvestments made in reliance on [the 5% permission in] [S]ection4(c)(6) must be essentially passive . . . [S]ection 4(c)(6) is not anunqualified grant of permission for a bank holding company toacquire or retain a 5 percent voting interest in any company. It is the

64. Thus, astute legal analysts' claims that law does not matter because banks in both Japan andAmerica face identical 5% limits on stock ownership are puzzling, since neither the stunting caused by theMcFadden Act nor the passivity rules of the Bank Holding Company Act are discussed. See, e.g., Coffee,supra note 37, at 1278-81, 1294-95.

65. Bank Holding Company Act of 1956, ch. 240, § 4(a), 70 Stat. 133 (1956) (codified at 12 U.S.C.§ 1843(a)(1) (1988)) (emphasis added). To be precise, an American bank cannot own any stock: thepermitted ownership is for a bank holding company. This distinction probably increases the costs of certaintransactions, but I assume that the effect is not large.

66. Id. § 4(c)(6), 12 U.S.C. § 1843(c)(6) (1988).67. Id. § 4(a), 12 U.S.C. § 1843(a)(1) (1988).68. See supra note 59 and accompanying text.

1952 [Vol. 102: 1927

Corporate Structure

Board's view that the prohibition against bank holding companies'engaging in nonbanking activities extends to joint ventures orconcerted action by a group of bank holding companies in anonbanking activity as entrepreneurs."'6 9

Congress did not, with the 5% rule, unleash banks to use small blocks of stockto gain influence in corporate boardrooms, or at least so ruled the FederalReserve. 70 Later the Board reiterated that it "believes that section 4(c)(6)should properly be interpreted as creating an exemption from the generalprohibitions.., on ownership .... only for passive investments amountingto not more than 5 percent .... 71

These explicit passivity interpretations were somewhat undercut by anearlier Federal Reserve Board interpretation concerning, ironically enough,keiretsu main bank cross-ownership. For a time the Bank Holding CompanyAct, if read literally, prohibited Japanese banks operating in the United Statesfrom owning stock in Japanese companies located in Japan. Although the Fednever enforced the extraterritorial reach of the Act, saying it was meant toapply to U.S. commerce, not Japanese commerce,72 the Board asked Congressto amend the Act when many foreign banks entered the United States in the1970's,73 and Congress did.74

These passivity rules could break down through regulatory reinterpretation,just as regulatory reinterpretation has expanded banks' securities power inrecent years, because the rulings come not from the clear command of thestatute, but from the Fed's plausible interpretation of that statute. Althoughthey may seem to be minutiae, they really are not, because they complete thebroad congressional policy preference exhibited in McFadden's branchingrestrictions, Glass-Steagall's stock ownership restrictions, and the BankHolding Company Act's line-of-business restrictions: to keep banks small, to

69. 4 F.R.R.S. 4-338.2, Jan. 22, 1986, available in LEXIS, Bankng library, FRRS file (Federal ReserveBoard ruling regarding whether activities of mutual insurance company are "closely related to banking").

70. Because the statute does not explicitly require passivity for a 5% blockholder and does notexplicitly cover the informal relationships in Japan, the words of the statute are ambiguous. In the FederalReserve Board's interpretation, if the main bank is not passive, it violates the Act, but a literalist mightinterpret the statute differently.

71. 12 C.F.R. § 225.137 (1990) (emphasis added). See generally PAULINE B. HELLER, FEDERAL BANKHOLDING COMPANY LAW § 4.03[2][a], at 4-60.9 (1992).

72. And then, since the Act did not exempt foreign banks' foreign operations from the Act's reach,the Fed did not adopt its subsequent passivity interpretations and said that activity, as long as there wasno control, passed muster. In re Dai-Ichi Kangyo Bank, 58 Fed. Res. Bull. 49, 49 (1972) ("[i]n light of the[Act's] purpose... to maintain separation of banking from commerce in the United States.") (emphasisadded).

73. INTERNATIONAL BANKING AcT OF 1978, S. REP. No. 1073, 95th Cong., 2d Sess. 1 (1978).Actually, the Federal Reserve Board sought Congressional action even before 1972. International BankingAct of 1976: Hearings Before the Subcomm. on Financial Institutions of the Senate Comm. on Banking,Housing and Urban Affairs, 94th Cong., 2d Sess. 21, 30 (1976) (statement of Fed Vice Chairman Gardnerquoting 1970 Senate testimony of Chairman Bums).

74. International Banking Act of 1978, Pub. L. 95-369, § 8(e), 92 Stat. 623 (1978) (codified at 12U.S.C. § 1841(h)(2) (1988)).

1993] 1953


keep private economic power unconcentrated, and to put a fault line betweenbanking and industry.

3. Main Banks and Universal Banks-The Potential for Control

If American banks tried to imitate either the Japanese main banks or theGerman universal banks, the Act would still deter them if the Fed revoked itspassivity rule and allowed activity short of control.75 Large investors want theoption to exert control, even if they rarely exercise it. When BerkshireHathaway buys a big block of stock and sits on a firm's board, it does notwant day-to-day control. It does, however, want the freedom to takecontrol-to replace the CEO in a crisis as it did in the recent SalomonBrothers scandal. Short of crisis, large shareholders with the potential to exertcontrol are more influential if not barred from control. Since crisis interventionis a key function of the foreign systems, a regulatory bar on control is a steepbarrier.

76

U.S. banks can and sometimes do assert control over firms, particularlysmall firms, 77 as lenders, without violating the Bank Holding Company Act.Banks whose holding companies own stock should be shy of doing so, becausethe source of control-debt or stock-would probably be unclear. Indeed, stockcan support the loan because in crisis stock can yield control by voting in newdirectors fasier than debt can yield control by enforcing covenants inbankruptcy. Moreover, even if U.S. banks can show that their control camefrom lending, not owning stock, the stock heightens the risk of equitablesubordination of their loans in bankruptcy and claims of lender liability outsidebankruptcy. Unlike main banks in Japan, which are subordinated by custom,informal agreement, and Ministry of Finance guidance, U.S. creditors canavoid equitable subordination (and lender liability) by inaction.78 The mainbank, already subordinated, can improve its return on its loan by fixing thefirm that is in crisis, not by ignoring the crisis.

75. When the literal reading of the statute prohibited Japanese banks from acquiring American banksif they were active in commerce in Japan, the Fed interpreted § 4(c)(6) as prohibiting only control, notactivity. See supra text accompanying note 72.

76. PAUL SHEARD, THE ROLE OF THE JAPANESE MAIN BANK WHEN BORROWING FIRMS ARE INFINANCIAL DISTRESS (Stanford Center for Economic Policy Research Working Paper No. 330, 1992).

77. Thus, I understand that some banks involved in leveraged buyouts have accepted the risks outlinedin this section.

78. See, e.g., Taylor v. Standard Gas & Electric Co., 306 U.S. 307 (1939); State National Bank of ElPaso v. Farah Mfg. Co., 678 S.W.2d 661 (Tex. Ct. App. 1984); J. Mark Ramseyer, Japanese Main Banksas a Regulatory Artifact: The Legal Framework (1991) (unpublished manuscript, on file with author).

1954 [Vol. 102: 1927

Corporate Structure

4. Main Banks and Universal Banks-The Suppression and Capture ofthe Securities Market

A society can move savings from households to firms through strongstockholding intermediaries or through a securities market. America's Glass-Steagall Act severed the intermediary channel from the securities channel,weakening American intermediaries by creating two channels. In Germany, thebanks captured the securities channel; the German securities market is weak,and the banks substantially control it, a result that allows for powerfulintermediaries, which can provide both banking and securities services to firmsand households. As we have seen, a big part of the German banks' votingpower comes from securities owned by the banks' customers and depositedwith the banks.

The Japanese situation is more complex. One might think that becauseJapan has a Glass-Steagall Act-imposed on it during the American postwaroccupation-it would resemble America in having two channels. One mightthen mistakenly argue that because Japan has a Glass-Steagall Act, Glass-Steagall did not restrain the development of powerful U.S. banks.

But such an argument misunderstands Japan's Glass-Steagall Act. In fact,until recently Japan has more resembled Germany in having essentially onechannel. Like the American version, the Japanese Act severed commercialbanks from investment banks. But the Japanese system then took a differentpath: Japanese postwar regulation skewed industry financing toward banks andaway from the securities market by (1) suppressing the bond market throughcollateralization and issuance regulations;79 (2) limiting competing sources of

corporate finance, such as equity issuance;8" (3) impeding the developmentof investment companies;8 ' (4) requiring that banks serve as trustees forbondholders when companies were allowed access to the bond market;82 and(5) holding down the interest rates paid on deposits to enable banks to profiteven when lending at low rates.

Properly understood, postwar Japan adopted two offsetting sets ofregulations. One set segmented finance but was not as severe as U.S.regulation-although Japanese banks could not issue securities, sell insuranceor own very large blocks of stock, they could become large and be active incorporate governance. The second set channeled finance through banks by

79. FRANCES MCCALL ROSENBLUTH, FINANCIAL POLITICS IN CONTEMPORARY JAPAN 157-66 (1989).80. ROBERT ZIELINSKI & NIGEL HOLLOWAY, UNEQUAL EQUITES: POWER AND RISK IN JAPAN'S

STOCK MARKET 156 (1991); Ramseyer, supra note 78, at 23-31; MIKUNI & Co., BANKING 5 (OccasionalPaper No. 2, 1987).

81. Hideki Kanda, Politics, Formalism, and the Elusive Goal of Investor Protection: Regulation ofStructured Investment Funds in Japan, 12 U. PA. INT'L L. REV. 569 (1991).

82. MASAHIKO AOKI, THE JAPANESE FIRM AS A SYSTEM OF A'TRIBUTES: A SURVEY AND RESEARCH

AGENDA 17-18 (Stanford Center for Economic Policy Research Publication No. 288, 1992). Thus Japanpushed corporate borrowers into commercial banks and limited savers' options.

19931 1955


ensuring that depositors had few options other than banks and that largecorporate borrowers had few nonbank financing sources.83 Thus, Japaneffectively "repealed" the American imposed Glass-Steagall separation, not byallowing commercial banking to mix with investment banking, but bystymieing a securities market and channeling savings and corporate financinginto the banking system. 84

Hence, although laws in both nations severed investment banking fromcommercial banking, the results differed, because, unlike the Americans, theJapanese allowed their banks nationwide operations, forced savers to use bankson terms favorable to the banks, and required corporations to seek financingthrough the banks. In operation, Japan did not have the Glass-Steagall Act thatAmerica had. America constructed two big competing financial channels; Japanchanneled both savings and finance through the banks. The single channelallowed the Japanese banks to be powerful enough to take a serious role incorporate ownership.

The Japanese single channel for finance is now weakening for two reasons.Success has given Japanese firms the luxury to retain earnings, thereby side-stepping the banks' control over the financing channel. (This success and theside-stepping it facilitated now make the banks' control over stock potentiallymore important than it has been.) Also, the channelling regulations that onceoffset Glass-Steagall are disappearing. Regulatory change has opened up thesecurities channel a bit, thereby weakening the commercial banks because theyhave not completely succeeded in getting regulators to allow them full entranceinto the newly-widening securities channel.85

E. Other Regulatory Impediments

Since my purpose here is not to catalog unending legal impediments butonly to show that the foreign systems would fall under American law, evenone show-stopper restriction-McFadden, for example-suffices. In addition,the other major U.S. financial intermediaries-insurers, mutual funds, andpensions-have historically also been precluded from taking big financialpositions in America's largest firms, preclusions consistent with the generalthesis that the structure of the large firm is highly sensitive to the structure of

83. Cf Ramseyer, supra note 78, at 23-31.84. See Bruce Kasman & Anthony P. Rodrigues, Financial Liberalization and Monetary Control in

Japan, FED. RESERVE BANK N.Y. Q. REv., Autumn 1991, at 28, 29-31 & n.4. Determining which is the"natural" base-American securities markets or Japanese banks-is difficult because America has burdenedbig banks while Japan has subsidized them. Nationwide banking seems a natural baseline, which onlyAmerica eliminated, thereby reducing nonsecurities alternatives for large firms and facilitating thedevelopment of a securities market. If large American banks had existed, they might have made large loansand stock investments that in America had to flow through the securities market.

85. This discussion of forcing funds through banks is offered as a comparison with Japan, not as apolicy recommendation.

1956 [Vol. 102: 1927

Corporate Structure

financial intermediaries, which in turn is highly sensitive to law. For half acentury, major American insurers were prohibited from owning any stock, 6

and mutual funds have been discouraged from acquiring influential blocks ofstock.87 Pension funds, while not formally prohibited from buying influentialblocks, are controlled by managers of large firms who discourage suchinfluence, rather than by managers of financial intermediaries who mightencourage it."

Securities laws historically have made communication among stockholderscostly. Until last year, ten stockholders who merely spoke with one anotherabout corporate events and managers risked violating proxy rules.8 9 Interbank(and interfinancier) communications among banks with large blocks of stockcould have been construed as a proxy solicitation, thus necessitating a publicfiling with the Securities and Exchange Commission. Under state antitakeoverlaws, group votes are generally sterilized, trigger poison pills, or violate"control share" statutes. 90

Why don't U.S. firms participate directly in each other's governance byholding large blocks of each other's stock? Although there are no explicitprohibitions, operating firms are poorly suited to hold large blocks of stockbecause they usually prefer to deploy their capital for other purposes. Even inJapan, where industrial cross-holdings play a role, financial intermediaries holdtwo-thirds of the blocks, while industrial firms hold only one-third of them.Moreover, American industrial cross-holdings would be taxed. In the 1930'sCongress passed a dividends received tax explicitly to discourage suchcorporate complexes. 9'

Why did the United States adopt so many impediments to institutionalvoice in corporate governance? American populism-a popular mistrust ofpowerful private financial intermediaries-and American interest groupjockeying yielded many of the restrictions. While German and Japanesepolitics display some similarities-particularly popular distrust in Germany andinterest group infighting in Japan-foreign politics has been different. Differentpolitical paths yielded different financial institutions, and different financialinstitutions yielded different corporate structures. We shall return to considerthese historical and political influences in Part IV. Before we do, however, we

86. Mark J. Roe, Foundations of Corporate Finance: The 1906 Pacification of the Insurance Industry,93 COLUM. L. REV. 639 (1993).

87. Mark J. Roe, Political Elements in the Creation of a Mutual Fund Industry, 139 U. PA. L. REV.1469, 1470 (1991).

88. See generally Roe, Modem Corporation, supra note 57.89. See Black, supra note 9, at 537-41. New proxy rules, adopted on October 15, 1992, reduce, but

do not eliminate, some of these possibilities.90. See, e.g., Cullen v. Milligan, 578 N.E.2d 123 (Ohio 1991) (applying Ohio control share statute);

Atlantis Group, Inc. v. Alizae Partners, No. 1:90 CV-937, 1991 WL 319384, at *1 (W.D. Mich. Aug. 27,1991) (applying Michigan control share statute).

91. Roe, supra note 87, at 1478. Today the tax rate is about 10% of the dividend. The tax alsodiscourages bank cross-holdings.

19931 1957


must first assess a preliminary economic argument-that financial evolutionnecessarily yields a fragmented financial system. After all, if fragmentation isinevitable, we need not bother with the difficult task of hypothesizing whetherintermediary participation in flattened corporate structures is on balanceadvantageous. Over time entropy will prevail, the argument goes, andownership will fragment.

III. FINANCIAL EVOLUTION IN GERMANY AND JAPAN?

In this Part, I briefly examine economic evolution here and in Germanyand Japan. While it is possible that financial evolution inevitably fragmentsownership of the large firm, the evidence suggests that intermediaries'economic evolution is uneven and that fragmentation is not economicallyinevitable. The financial markets in Germany and Japan may not be movingtoward full American-style securities markets with fragmented ownership; thereis as much evidence of stability and concentration abroad as there is offragmentation. It is equally plausible that our markets and ownership forms areapproaching-weakly and unevenly-the more concentrated structures ofGermany and Japan. Thus, we should re-interpret recent trends in the UnitedStates toward ownership concentration and institutional voice as the delayedresults of the repeated historical suppression of powerful American financialintermediaries.

A. The Evolutionary Argument

Do Germany and Japan lag behind America's financial evolution? In time,the argument runs, finance liquifies and disintermediates. Securitization,diversification, and fragmentation of stockholdings erode big institutionalblocks. Powerful intermediaries wither as stock disperses into small blocksheld by individuals and weak institutions. Commercial paper replaces bankloans, eroding banks' interest in holding stock in their former loan customers.Hence, the universal problems of financial organization eventually compeldispersed ownership, managerial agency problems, and structures for dealingwith agency costs similar to those in America. According to the evolutionaryargument, then, the powerful financial institutions in Germany and Japan areonly temporary and will ultimately come to resemble U.S. financialinstitutions, which represent the highest form of financial development.Inevitable economic forces will erode the presence of institutions at the top offoreign firms.

I question this view. True, German banks' boardroom power is understress.92 And the Japanese main bank system is under severe stress from

92. See infra notes 123-129 and accompanying text.

1958 [Vol. 102: 1927

Corporate Structure

increased securitization of debt in Japan, the internationalization of financialmarkets, Bank of International Settlements capital standards that penalize stockownership, the increase in Japanese corporate retained earnings, and a sinkingJapanese stock market. I suspect that key features of the main banksystem-those involving bank control of large firm new financing-will notsurvive and may already be gone. Nevertheless, German banks' big stockholdings have increased, not decreased, over the last twenty-five years, andJapanese bank ownership of large Japanese firms has been rock-solid atroughly 20% of the outstanding shares for the past twenty-five years. Thesetrends belie the inevitability of fragmentation. And, as I discuss in Part IV, ifthere is financial evolution, political forces are playing a role.

B. Problems with the Evolutionary Argument

Securitization and economic change are indisputable facts. Their effects,and the possibility of substitutes, are in doubt.

1. Securitization

Securitization produces three threats to German universal banks andJapanese main banks. First, as debt securitizes, commercial paper and bondsreplace bank loans, so commercial banks' influence as lender declines. Second,as banks make fewer loans to their customers, they could become lessinterested in holding their customers' stock; if they sell the stock, commercialbanks' influence as stockholder declines. Third, as liquid securities marketsdevelop, institutions fragment their stock holdings and, the argument wouldrun, no financial institution, and certainly no bank, would want to hold a bigblock of stock; institutional influence through big blockholding declines.

The first threat is real. As some debt has securitized, banks' influence hasdeclined. But must banks' power from stock wither as debt increasinglysecuritizes?" Does the realization of the first threat to bank power---decliningloans-demand realization of the second threat-declining stock? The evidencesuggests not. As Table VI shows, German bank ownership of large blocks ofstock in the largest firms has increased during the past twenty years, and, asseen in Table I, Japanese bank ownership of stock in large firms was stablethroughout the 1980's. 94 Those predicting fragmentation may well have

93. Banks are a conduit between savers and borrowers. Depositors place money in banks, and thebanks then lend those funds, but are obligated to repay the deposits even if the loans are not repaid. Yetsecurities markets are increasingly assuming these functions. Depositors who used to deposit their moneyin the bank now lend the money to corporations without using banks as intermediaries. To bypass thebanks, providers of short-term funds buy commercial paper from industrial corporations and providers oflong-term loans buy bonds.

94. Between 1977 and 1992, the banks sold a modest 1% when Japan lowered the maximum bankholding from 10% to 5% of the outstanding shares of stock, to take effect in stages through 1987. The sell-

19591993]


conflated an inevitable weakening of fractional reserve banking-the firstthreat-with a weakening of large stockholding institutions. The first is likelyand the current trend the world over; the second may not at all be likely andis certainly not yet supported by the German and Japanese evidence.

TABLE VI. German Banks 5% or Greater Positionsin the 100 Largest Firms95

Year 1972 1975 1978 1980 1982 1984 1986 1988 1990

Number of Positions 20 21 25 26 25 29 30 30 33

A common misconception is that German bank influence comes only fromcredit. Americans are so inured to the separation of banks from commerce thatwe tend to view credit allocation as the banks' primary means of influence.Although German banks' absolute power over credit has already withered forthe large German firms, 9 6 their stock-based control over the proxy system has

off stabilized in 1985. This 1% sell-off equals the bank-owned stock above 5% in 1977, as Table XIIshows. Law and Japanese politics, not securitization, best explain the sell-off. Japanese Federal TradeCommission, The Actual Conditions of the Six Major Corporate Groups 2 (Aug. 1 1989) (unpublishedreport, on file with author).

Now that Japanese stock prices are depressed, another sell-off has been predicted (but not observed).This situation mirrors the one in which stock prices were high and a sell-off was also predicted. See LarryZoglin, Stable Cross-Holding of Shares Likely to Withstand Pressures, JAPAN ECON. J., June 21, 1986, at7. Tables III and XI show that a sell-off did not occur. In general, though, a sell-off at high prices seemsmore plausible then one at low prices. When prices are high, insiders reap a gain by selling, but when theyare low, insiders incur a loss.

95. Source: JORGEN BOHM, DER EINFLUSS DER BANKEN AUF GROSSUNTERNEHMEN 238 (1992). Noris this a repositioning of stock. Banks' direct ownership of public companies has been rising:

TABLE VII. German Banks' Ownership of Public Companies

Year 1960 1970 1980 1985 1988 1990

Percentage of Public CompanyStock Owned by Banks 6% 7% 9% 8% 9% 10%

Source: Julian Franks & Colin Mayer, Corporate Control: A Synthesis of the International Evidence 24(Nov. 1992) (unpublished manuscript, on file with author) (using Deutsche Bank data).

The reader will recognize a discrepancy between Table VI and Table VIII, which shows twenty-two5% blocks. Table VI seems to include blocks in the top German banks, but Table VII does not. The keyissue here is that the number of big block positions is increasing not decreasing.

96. Michael H. Best & Jane Humphries, The City and Industrial Decline, in THE DECLINE OF THEBRITISH ECONOMY 223, 224 (Bernard Elbaum & William Lazonick eds., 1986). Since 1970, retainedearnings constitute 84% of new German corporate financing-about the same level as in America. CLAUDIOBORIO, LEVERAGE AND FINANCING OF NON-FINANCIAL COMPANIES: AN INTERNATIONAL PERSPECTIVE, 15(Bank for International Settlements Economic Papers, No. 27, 1990).

Of the (small) amount of capital projects that are financed externally (i.e., without relying on retainedearnings), in Germany, 70% is provided by the banks while in America only 25% is provided by the banks.Randall J. Pozdena, Commerce and Banking: The German Case, WEEKLY LETTER (Fed. Reserve Boardof S.F., Cal.) Dec. 18, 1987, at 2; see also J.S.S. EDWARDS & KLAUS FISCHER, BANKS, FINANCE AND

1960

Corporate Structure

not, and their big blockholdings are increasing.97 True, since Japanese credit

control has weakened only in recent years, we cannot be sure Japan will take

the same path as Germany did. Because bank financing is higher in Japan than

in Germany, the banks' influence as lender-whose decline constitutes the first

threat-will decrease as competing financing channels open. But if Japanesebanks can enter the securities channel, the first threat may not be devastating;

such entry would not only enable them to keep influence via debt, albeit

securitized debt, but would enable them to retain a weakened credit-basedmotivation to hold stock.

The third threat is that securitization would necessarily destroy big blocks.

In theory, this is false.98 Debt and stock may disintermediate from direct

ownership by banks and reappear in a complex intermediary that combines

mutual funds, insurance funds, brokerage stock, and pensions. To some extent,

this combination already exists in Germany. Deutsche Bank combines several

of these, and the new Dresdner-Allianz Insurance combination combinesseveral others.99 Stock-based power seems to be increasing in Japan, as a

substitute for credit-based power. Japanese banks now dispatch more directors

to industry, not fewer,00 and Japanese insurers are becoming more

assertive.'0 ' Furthermore, the extent of the disintegration depends upon

Japan's Glass-Steagall Act and the absence of strong nonbank intermediaries.Without Glass-Steagall, Japanese banks might become brokers or sponsor

mutual funds; disintermediated debt and stock could consolidate and be held

by brokers and mutual funds affiliated with banks. 02 In fact, commercial

INVESTMENT IN WEST GERMANY SINCE 1970 21 (Centre for Economic Policy Research Discussion Paper

Series No. 497, 1991) (arguing that German stock corporations "relied hardly at all on bank borrowing as

a source of finance for investment over the period 1971-85, and instead were largely internally financed").97. EDWARDS & FISCHER, supra note 96.98. The evolutionary theorists mistakenly offer this syllogism:

(1) Banks are the intermediaries with governance influence in Germany and Japan.(2) Banks' core function is to lend money.(3) Securitization and retained earnings marginalize banks' credit.(4) Therefore, as securitization reduces bankers control over credit, financial institutions' influencemust weaken.

Statements (1), (2), and (3) are true, but, due to substitutes and stock-based power, (4) is not.99. See infra note 115 and accompanying text.100. Japanese Federal Trade Commission, supra note 94, at 2; cf. Kaplan & Minton, supra note 36,

at 20 (noting more bank interventions as stock performance weakens). Whether bank monitoring should

be interpreted as the intervention of residual stockholder, or of a large creditor, or as intervention arising

from contractual relations in the production process, is discussed briefly infra notes 194-198 and

accompanying text and in detail in Ronald J. Gilson & Mark J. Roe, Understanding the Japanese Keiretsu:

Overlaps Between Corporate Governance and Industrial Organization, 102 YALE L.J. 871 (1993).101. KESTER, supra note 31, at 216 (Japan's largest insurer, Nippon Life, says it used to be passive

as a shareholder but is now beginning to assert its position as an investor where it affects the return on

investment). Although insurers have increased their stockholdings, partly to pick up the stock that bankswere forced to sell when the lid was lowered from 10% to 5%, they are now reaching the limits of their

legal (and perhaps financial) ability to hold stock. JACK MCDONALD, ORIGINS AND IMPLICATIONS OF

CROSS-HOLDINGS IN JAPANESE COMPANIES 9 (Graduate School of Business, Stanford University TechnicalNote No. 79, 1991).

102.) Even if bond debt replaces bank debt, if bank affiliates hold that debt, elements of the main bank

system would be recreated, although in muted form. Indeed, as the Japanese Glass-Steagall Act has been

196119931


bank evolution to a more complex intermediary is budding in Japan, and theMinistry of Finance has proposed allowing banks to buy, sell, trade, and formmutual funds for commercial paper and bonds. Investment companies andfunded pensions are expected to flourish.10 3 Hence, Japanese banks' powerin corporate governance will continue to erode only if they lack the politicalpower to get Japan's version of Glass-Steagall rolled back. To date they havebeen only partly successful.' °4

Even if German and Japanese banks sold their stock, someone would ownthe stock that banks now own. Dispersed individuals might hold the stock insmall blocks, although that is unlikely; small institutions might buy smallblocks of the stock, and that is possible. Or, new concentrations might arise,such as larger German banker-broker networks, or powerful mutual funds, orexpanded versions of the Japanese system of customer-supplier cross-ownership. Finally, the Japanese banks might "sell" their stock to affiliates, ifthe banks induce repeal of the Glass-Steagall Act.'05

2. Autonomy

The evolutionary paradigm for Germany and Japan posits that powerfulintermediaries develop industry, and their power then erodes due to thesecurities market's disintegrating effects. If there were powerful intermediariesresembling the universal banks or the main banks in America's past, and theyweakened as they were displaced by a superior securities market, then theargument that foreign nations now lag behind America would be moreconvincing. We could predict a single path for financial fragmentation. Butpowerful stockholding intermediaries were generally absent in Americanhistory. We should not view the evolution of financial systems as following asingle economic path, in which every system ultimately converges on highlydiversified portfolios, fragmented ownership, and weak intermediaries.

Details from American history are suggestive. In nineteenth-century NewEngland, entrepreneurs tightly bound their operating firms with banks.Similarly, the Japanese zaibatsu families bound their industrial firms withbanks, and Deutsche Bank in its early history helped finance new companieslike Siemens and AEG. The foreign banks grew into big national financial

rolled back slightly, the Japanese banks have moved into the debt securities market. MASAHIKO AOKI,INFORMATION, INCENTIVES, AND BARGAINING IN THE JAPANESE ECONOMY 141 (1988).

103. ROSENBLUTH, supra note 79, at 75; Kanda, supra note 81, at 569.104. Japan Will Allow Overlap of Banks, Brokerage Firms, WALL ST. J., Dec. 18, 1992, at A4

(describing slow, partial erosion of Japanese Glass-Steagall Act).105. The result is likely to depend on the future nature of business. The Japanese managerial style,

labor relations, and rapidly changing product cycle have favored today's ownership structure. If theseunderlying business features change, the old structures may become less useful. Cf. Masahiko Aoki, Towardan Economic Model of the Japanese Firm, 28 J. ECON. LIT. 1 (1990). A plausible prediction is that somefirms, like those whose business is easily understood by investors in the securities market, will securitizewhile others will not.

1962 [Vol. 102: 1927

Corporate Structure

institutions, the American banks did not, and the ties between the Americanentrepreneurs and their local banks withered. Why? As economic opportunitiesshifted from New England to the national economy, the New England bankscould not get good information about distant firms, and the bankers couldparticipate in the national economy only as passive buyers of short-termcommercial paper.10 6 "[T]he legal prohibitions against branch banking andthe distance between economic centres produced... relatively small, not veryclosely connected, short-term markets. . . .Because of the prohibition on

interstate branching, a national market had to await the development of a...commercial paper [market]."' 17 Entrepreneurs affiliated with banks could gonational, as economic opportunities certainly did, but the bankers, due to

interstate branching restrictions, could not.'"Recent German trends and Japanese economic history also undercut the

argument that their ownership structures must evolve toward the Americanmodel. First, German managers have behaved inconsistently toward bankercontrol of the proxy system. Many German companies have limited the votingof outsiders, not to undercut the power of bankers, but to thwart hostiletakeovers.' 09 Managers, it seems, choose to have large, sometimes influential(but not dominating) blocks of stock controlled by financial intermediarieswhen they fear a hostile takeover. Second, Japanese cross-holdings are a recent'phenomenon. Before World War II, the zaibatsu combined banks and industrialfirms, under family ownership, with banks as the weaker player. During World

War II, the Japanese government munitions planners weakened the authorityof the family owners. After World War II, the American occupation broke upthe zaibatsu and distributed stock widely. Eventually enough stock fell intobank hands to give rise to banker authority. During the 1960's cross-holdings

accelerated for two reasons. Many Japanese managers feared that joint ventureswith American companies would end up with the American firm owning theJapanese firm; they wanted a lot of stock placed in safe, friendly Japanese

hands. In addition, in the late 1960's corporate Japan feared that the Japanese

106. Naomi R. Lamoreaux, Information Problems and Banks' Specialization in Short-Term

Commercial Lending: New England in the Nineteenth Century, in INSIDE THE BUSINESS ENTERPRISE:HISTORICAL PERSPECTIVES ON THE USE OF INFORMATION 161, 180 (Peter Temin ed., 1991) ("[N]ow that

firms could issue their IOUs through note brokers, who would market them to banks and financial

intermediaries across the country, banks lost their ability to assess a customer's total indebtedness.")(emphasis added). Since banks have an advantage in information gathering, one wonders why bankersceded the profits to investment brokers, other than because, unlike banks, brokers could market notesthroughout the country.

107. Lance E. Davis, The Capital Markets and Industrial Concentration: The U.S. and U.K., A

Comparative Study, 19 ECON. HIST. REV. (2d ser.) 255, 260 (1966).108. Massachusetts (like other states) also prohibited its banks from lending to out-of-state firms

(although it apparently didn't stop them from buying securitized commercial paper). Lance E. Davis,

Capital Immobilities and Finance Capitalism: A Study of Economic Evolution in the United States 1820-

1920, 1 EXPLORATIONS IN ENTREPRENEURIAL HIST. 88, 99 (1963). Regulation sufficiently explains theresult; whether it was necessary, we cannot know.

109. BAUMS, supra note 16, at 10.

19631993]


government would sell stock from financially-distressed brokerage houses; thisrisk and depressed stock prices created fear of American-style takeovers, whichcross-holdings, including holdings by banks, would deter. The cross-holdingsthat are at the core of modem Japanese corporate governance are a post-WorldWar II institution.11t The bank-centered keiretsu succeeded the old-linezaibatsu. Then, during the postwar decades, newly rising big firms independentof the traditional keiretsu have chosen the same concentrated institutionalownership. '

American managers might have encouraged strong financial intermediariesto enter the boardroom had they feared similar uncertainties. If Americanmanagers had not been so successful in defeating takeovers with poison pills,legislative antitakeover barriers, and the luck of a rising stock market, theymight have invited intermediaries into corporate boardrooms. They may wellhave disliked powerful stockholders but probably detested hostile takeoversmore. Managers and financiers could have formed an irrepressible politicalcoalition to undo existing American legal barriers to a close nexus betweenfinance and industry.11 2

Although I cannot rely on this "might-have-been" claim to prove thatevolution into fragmented ownership and liquid securities markets is notinevitable, I can point to a few targets that sought big blockholders to protectthemselves from takeovers." 3 Had U.S. managers not defeated takeovers inother ways, big blocks might have been management's next defense. Similarly,after the prospect of American-style takeovers arose in Germany,"4 DresdnerBank, Allianz Insurance Company, and Hoechst, a huge chemical firm,developed major cross-shareholdings among themselves and other German

I10. McDONALD, supra note 101, at 1.111. See Table XI (some of the 14 largest firms, shown in Table III, overlap with the new

"independents" shown in Table XI). Some see the new Japanese large firms that have arisen outside theold-line successors to the zaibatsu as a sign of institutional ownership's erosion. Coffee, supra note 37, at1301-02. Since, however, the new firms' institutional ownership structure is identical to that of the old-linefirms, their rise does not show a decay in institutional ownership. Moreover, because the new firms are atad more profitable than the old ones, some conclude that cross-ownership is deleterious to shareholders.But since the new firms have the same ownership structure (and frequently have founding families withsome stock as well), institutional ownership structure is not a relevant variable.

112. Cf Carol J. Loomis, The New J.P. Morgans, TIME, Feb. 29, 1988, at 44 (listing LBO's withfinanciers on board). Many LBO's dealt with managers' misdirected use of free cash flow to build empires,a problem Japanese firms are only now beginning to face. Thus, the point here is not that LBO's andinstitutional ownership are perfect equivalents, but that LBO's flatten authority by bringing financiers intothe boardroom, thus calling into question the notion that economic evolution inevitably drives large firmstoward fragmented ownership.

113. Vineeta Anand, Warren Buffett Effect: A Quick Jump in Stock Prices, Aug. 23, 1991, at 8, 8 ("InSeptember 1987, Buffett infused $700 million into Salomon, then facing a takeover threat from RonaldPerelman. He was rewarded with two seats on the board . .. "); David A. Vise, CBS Loses $114 Millionin Quarter A Record, WASH. POST, Nov. 13, 1985, at El, El ("Loews Corp. Chairman Laurence A. Tischwas elected to the CBS board ... [Loews' stock ownership] is expected to bring stability to a companythat has been the subject of intense takeover speculation ... ").

114. The most prominent of these was Pirelli's multiyear siege of Continental, the large German tiremanufacturer. See Andrew Joncus, Continental AG Shareholders Deliver Pirelli Another Blow, WALL ST.J. EUR., July 6, 1992, at 4.

1964 [Vol. 102: 1927

Corporate Structure

industrial companies." 5 The Japanese construction after World War II of

new vertical keiretsu around the firms that emerged independent of the old-linekeiretsu and the acceleration of cross-holdings in the 1960's are similar. Thus,features of all three nations-the recent German keiretsu-style cross-ownership,the fact that Japanese cross-ownership is a post-World War II institution, and

the scattered American antitakeover blocks-all tend to contradict theinevitability of fragmentation.

3. American Evolution

Commentators assert that Germany and Japan will ultimately evolve to

American-style diffuse ownership. Perhaps they will. Japanese main banks are

undergreat financial pressure, which could lead them to sell stock to raise

cash; without affiliated allies such as bank-controlled trusts, mutual funds or

pensions to relieve some of the pressure and hold some stock, it is difficult to

see how the Japanese system can continue in its present form. Yet, unnoticed

is the prospect that U.S. firms may be evolving-weakly and

unevenly-toward the German and Japanese style of ownership. U.S.

intermediaries have steadily taken on new functions. The next natural

evolutionary stage in American corporate finance might be the entrance of

financial institutions into corporate boardrooms,'"6 perhaps by electing

directors whose loyalty runs to them, not the managers. Indeed, Berkshire

Hathaway, an insurer whose big blocks and legal authority to buy them are

recent acquisitions, already vaguely resembles the German universal banks and

Japanese main banks. And, leveraged buyouts, although different in key ways,

also increase the voice of institutions in corporate governance."17

True, the five largest American holders rarely own much more than 5%

of the twenty-five largest American firms, and only a few approach the level

of German ownership and voting and of Japanese concentration." 8

Nevertheless, some institutions have been active in corporategovernance-making shareholder proposals, seeking to elect directors, and

115. Hans Otto Eglau, Allianz/Dresdner Bank-Vermachtet und Verschachtelt [Empowered and

Interconnected], DIE ZErr, Aug. 16, 1991, at 19. The holdings were also tax-driven. Dresdner is Germany's

second-largest bank, Allianz its largest insurer, and Hoechst the sixth-largest industrial firm. Lufthansa and

Bayer are expected to join. The cartoon accompanying this article suggests popular reproach to the

concentration of private economic power by showing an Allianz-Dresdner octopus gobbling up German

industry. Id. Banks are hubs of new industrial networks elsewhere in Germany, with cross-representation

of network members on boards. Kirsten S. Wever & Christopher S. Allen, Is Germany a Model for

Managers?, HARV. BUS. REV., Sept.-Oct. 1992, at 36, 42.116. Cf. Robert C. Clark, The Four Stages of Capitalism, 94 HARV. L. REV. 561 (1981).117. Gilson & Roe, supra note 100, at 902-04 (discussing similarities and differences); Jensen, supra

note 9, at 65-66 (noting similarities).118. The rarities are mainly the large blocks held by Berkshire Hathaway. Coca-Cola had a five-

shareholder concentration level of 20%, as might a Japanese firm, but due to legal anomalies-holdingsby Berkshire Hathaway, an insurance holding company operating outside the norm, and a bank exemptfrom the Bank Holding Company Act's general provisions.

196519931


petitioning the Securities and Exchange Commission to loosen restraints ontheir activity."9 If these trends continue and expand to include holding bigblocks of stock, U.S. corporate ownership may someday begin to resemble theGerman and Japanese systems.

American aggregate concentration is already a pale imitation of foreigncorporate structures. The largest twenty-five institutional investors vote, onaverage, 16% of the stock of the twenty-five largest Americancorporations." 0 While this is a far cry from the three institutions in Germanythat vote 40%, or the five in Japan that vote 20%, it shows that U.S. large firmownership is no longer that of an atomized Berle-Means corporation. Andshareholder concentration for the next tier of U.S. firms is even greater,approaching the level of concentration in Japan. Since many of these second-tier stockholders are mutual funds and pensions, we can see that the U.S. trendover the past century has been, first, a suppression early in the century of the"basic" intermediaries (banks and insurers), followed by an emergence in the1990's of other institutions (mutual funds and pensions) as substitutes.

4. Economic Evolution and the Cost of Capital

A meta-economic explanation, atop the classical one, helps explain thedifferent laws that influence corporate structures. Societies afflicted withcapital scarcity have reason to organize firms differently than societies withmore abundant capital. Historically, capital and natural resources wererelatively abundant in the United States. In Germany and Japan, on the otherhand, capital was scarce in the post-World War I1 decades.12' Whenever aresource is relatively scarce, people and institutions are willing to pay tomonitor its proper use. In the American southwest, water is scarce andmetered, and its use is a serious legal issue; elsewhere, where water isabundant, its use is unmetered and unregulated. Likewise, when capital becamerelatively scarcer in the United States, as reflected in the rising interest ratesof the 1980's, more mechanisms to monitor its deployment came forth; andwhen Japanese success led to relatively greater stores of capital in the 1990's,monitoring mechanisms may have declined. 22

119. E.g., Randall Smith, Calpers Mulls Stakes in Funds Seeking Changes in Firns' StrategyGovernance, WALL ST. J., Dec. 31, 1992, at A3.

120. Brancato et al., supra note 10, at tbl. 7.121. Although the cost of capital that Japan "posted"--its interest rate-have made its capital seem

cheap, in fact bank rationing effectively reflected a capital scarcity after World War II.122. This fact could revive the evolutionary argument; one could argue that as economic tasks change,

the level of monitoring changes. For example, German and Japanese corporate structures may beparticularly adept at managing new investments and thus may be most suitable during a sustained generaleconomic expansion. But in mature economies some firms expand while some contract, making a mixedsystem suitable.

1966 [Vol. 102: 1927

Corporate Structure

Thus, as it turns out, current data indicate that fragmentation is not yeteconomically inevitable; there is as much (or more) evolutionary concentrationhere as there is evolutionary fragmentation there.

IV. POLITICAL EVOLUTION IN GERMANY AND JAPAN?

Even if German and Japanese stock ownership were fragmenting (andAmerican ownership were not concentrating), the incompleteness of currentcorporate theories would persist. We would need to determine the degree towhich politics in Germany and Japan is inducing financial evolution in thosecountries. Financial fragmentation could be inherent in twentieth-centurydemocracy, rather than merely inherent in American democracy as Ihypothesized earlier. Both foreign nations' political and corporate historiescould have evolved together: large industry emerged abroad whennondemocratic governments kept fragmenting forces in check; indeed, today'sforeign democracies affect corporate structures, but in ways and degrees

different from that which occurred in the American past.Although my goal here is not to provide a definitive history of how

German and Japanese politics affected their corporate structures- I commendthat task to those with the requisite knowledge and language skills, which Ilack-even a cursory inquiry by a nonexpert shows that political forces are atwork in Germany and Japan, shaping the governance of the large firm. Indeed,how could it be otherwise? The movement of capital from savers to firms hasto attract political attention and affect either public opinion or interest groups,whose interactions with political institutions will influence the organization offinancial institutions, which in turn will affect the ownership structure of thelarge firm.

A. German "Populism"

Deutsche Bank reviews whether to allow its employees to chair anindustrial firm's board, and some German banks have sold directly-held stock.

Is this part of an economic evolution? Perhaps so, but the German banks areunder intense political pressure 3 from forces not unlike those in the UnitedStates that historically disabled powerful financial intermediaries. German

123. John Dornberg, The Spreading Might of Deutsche Bank, N.Y. TIMES, Sept. 23, 1990, at 28(reporting impulse among Social Democrats and market-oriented Free Democrats to reduce banks' power);

Ferdinand Protzman, Mighty German Banks Face Curb, N.Y. TIMES, Nov. 7, 1989, at Dl, DI (reporting

that "parliament has begun studying steps to limit the banks' shareholdings, their seats on boards and theirinfluence in corporate decision-making."); Terence Roth, West German Banks Face Threat of ReducedInfluence in Industry: Bonn Wll Consider Rules to Curb Their Holdings and Seats in Boardrooms, WALL

ST. J., July 18, 1989, at A20, A20 (reporting that "[w]ith main stream politics coming into play, bankersworry that they'll be forced to sell parts of their sizable equity holdings in West German industry, thusthreatening their dominant position in the country's equity markets.").

19671993]


politicians claim that the bankers' power clashes with the German "socialorder."'124 Parliamentary reports attack the banks. Members of Parliamentdesire to limit "the percentage of equity a bank could maintain in a nonbankenterprise and cut the number of supervisory board positions a bank executivecould hold,"'l and look to the American Bank Holding Company Act forguidance.126 Managers are said to want to get bankers out of supervisoryboards, although they hesitate to say so publicly.27 Executives in Germany'smid-sized firms oppose bankers' influence and want laws prohibiting Germanbanks from owning nonbank stock. 2 8 Fear of powerful banks "form[s] thebackdrop of many economic and political discussions. Conducted in the press,on radio, and on television, among scholars, and, most importantly, politicalgroups, these discussions use emotion-laden terms such as 'bankocracy,''dominion of finance capital,' and similar verbal symbols."'29

The banks have bowed to this political pressure, at least verbally.130

After German politicians lit a storm of political protest in the region where abank-propelled takeover target operated, the bankers lowered their publicprofile. 131 Moreover, Deutsche Bank, Germany's largest bank, has stated that

124. Hans-Jacob Kriimmel, German Universal Banking Scrutinized: Some Remarks Concerning theGessler Report, 4 J. BANKING & FIN. 33 (1980); see also Der Herr des Geldes [The Money Man], DERSPIEGEL, Mar. 13, 1989, at 20; Die Geheimrdte der Nation [The Nation's Secret Council],INDUSTRIEMAGAZIN, Apr. 1987, at 27; Horst Greiffenberg, Die Macht der Banken [The Power of theBanks], VERBRAUCHERPOLITISCHE HEFrE, Dec. 1987, at 85; Jorg Huffschmid, Demokratische AlternativenderBankpolitik [DemocraticAlternativesfor the Politics ofBanking], VERBRACHERPOLITISCHE HEF'E, Dec.1987, at 111; Zwischen Bonn und Banken: Finanzdiplomat Hermann Abs [Between Bonn and the Banks:Financial Diplomat Hermann Abs], DER SPIEGEL, Nov. 3, 1965, at 10.

125. Dornberg, supra note 123, at 28.126. Wolfram Eckstein, The Role of the Banks in Corporate Concentration in West Germany, 136

ZEITSCHRiFT FOR DIE GESAMTE STAATSWISSENSCHAFr [J. INSTITUTIONAL & THEORETICAL EcON.] 467,480(1980) (Eckstein was Secretary General of the Monopoly Commission); see also Johannes Kdndgen, Dutiesof Banks in Voting Their Clients' Stock I 1 (July 1992) (unpublished conference paper, on file with author)(discussing parliamentary hearings of the committee for economic affairs).

127. Otto Graf Lambsdorff, Das Machtgeflecht der Banken Lichten, FRANKFURTER ALLGE.IEINEZEITUNG, Aug. 22, 1989, at 10 (leading German politician says senior managers privately tell him that theywish to see banker power reduced). When takeovers became a real possibility, however, German managersreconsidered.

128. Bundersverband mittelstlndischer Wirtschaf, Expose zur "Macht der Banken" 9 (Apr. 24, 1991)(unpublished results of survey conducted by German association of mid-sized businesses, on file withauthor). Eighty-eight percent of the businesses favor restricting the banks. Obviously mid-sized businessopposition could reflect interest group opposition to banker power.

129. H.E. Btischgen, The Universal Banking System in the Federal Republic of Germany, 2 J. Co\1P.CORP. L. & SEC. REG. 1, 25 (1979).

130. Krtimmel, supra note 124, at 53 (stating that bankers "continue with their traditional tendencyto elude public controversy by a flexible attitude and not to rise against the zeitgeist, even if they areconvinced they have the better arguments"); Die FDP will die Mineraldlsteuer erh~ihen stdrken,FRANKFURTER ALLOEMEINE ZEITUNG, Aug. 19, 1989, at II (leading German politicians form workinggroup to curb banker power and some expect the banks to understand that prudent self-limitation isnecessary); Fusion doch ohne Beschluss zur Bankenmacht, FRANKFURTER ALLEGEMEINE ZEITUNG, Sept.7, 1989 at 17-18.

131. "Once the [target's] employees protested the [proposed] deal, the politicians in southem Germany,where [the target] is based, began to send up a hue and cry [which caused Deutsche Bank to change itsposition and opposed the merger.]" Jackey Gold, M&A Continental Style, FIN. WORLD, Mar. 5, 1991, at37, 37.

1968 [Vol. 102: 1927

Corporate Structure

the political costs are too great to maintain high visibility in corporategovernance--either as an owner or as a director.132 To dampen publicprotest, which German bankers may believe managers could use to getAmerican-style legal restrictions, German banks have lowered their publicprofile and announced that they will not fight curbs on their control over theproxy machinery.133 While German banks do not seem to be withdrawingfrom stock ownership yet, they may well be starting to use their stock lessaggressively. Political pressures, not just economic evolution, account for someof this laxness.

True, this German "populism" has historically been weaker than theAmerican strain: "Germany has never known anything like the fear andresentment that monopoly used to arouse in the United States. . . .Many

Germans find it difficult to believe that something growing up without orderand control, like a competitive market, could not be improved by applying alittle discipline."' 4 In America, antibank sentiments combined with powerfulinterest groups-small bankers or managers, for example-to produce lawsrestricting large banks, giving deposit insurance to small banks, and protectingmanagers from takeovers. The restraints on German banks, in contrast, areinformal and self-imposed, designed to avoid formal restrictions. Why doesn'tGermany enact the formal restrictions anyway, if the popular sentiment and theinterest group motivations are there? Relative weakness of the sentiment andmotive is only part of the answer. Political structure and different formal limitsare also important. The German political structure dilutes the political effectof both anti-big-bank popular opinion and anti-big-bank interest groups.German citizens vote for a party which, based on its percentage of the nationalvote, gets a proportionate number of Parliamentary seats.'35 The party ismore important than the candidate. Local bankers and managers (and theircampaign contributions) play a less important role in German elections thanthey do in U.S. congressional elections. 136

132. Role of the Financial Services Sector: Hearings Before the Task Force on the InternationalCompetitiveness of U.S. Financial Institutions of the Subcomm. on Financial Institutions Supervision,Regulation and Insurance of the House Comm. on Banking, Finance and Urban Affairs, 101st Cong., 2dSess. 164-65 (1990). As of 1990, this statement is reflected more in Deutsche Bank's rhetoric than its

action, insofar as concerns its ownership blocks in Germany's 100 largest firms, which have beenincreasing, not decreasing. See Table VI.

133. Friedrich K. Kiibler, Institutional Owners and Corporate Managers: A German Dilemma, 57

BROOK. L. REV. 97 (1991). American institutional restraint is similar. California's huge state pension fundapparently fears that "an organized Calpers-backed attempt to force management changes could trigger areaction that could curb the big fund's freedom to act independently," in a fashion similar to the restraintsarising from the 1980's takeovers. Smith, supra note 119, at A3.

134. HENRY C. WALLICH, MAINSPRINGS OF THE GERMAN REVIVAL 136-37 (1955).

135. LEWIS J. EDINGER, WEST GERMAN POLITICS 119-20, 149, 172 & n.6 (1986).

136. Id. at 148 (stating that power of political parties "discourag[es] interest associations fromsupporting independent deputies and [strengthens] cohesion within the parliamentary parties. The leaders... can use the threat of expulsion to keep dissidents in line...").

196919931


Moreover, popular opinion and interest groups do formally shape theGerman boardroom and dilute the power of the German banks, throughcodetermination. It injects employees-white collar, blue collar, and union-represented-into the boardrooms of the largest German firms, now givingthem half of the supervisory board's seats not due to private contract-makingor purely economic evolution, but due to the way the German Parliamentsettled political conflict. Codetermination is a counterweight to capital-thehistorical origins of codetermination are rooted in the German Parliament'sefforts to co-opt revolutionary forces after the German revolution of 1918.137Parliament expanded codetermination thereafter, most recently in 1976 whenit sought to pacify unions after a wave of strikes in the 1970's.' While Iam unaware of a probing political analysis of Germany's expansion ofcodetermination in 1976, I doubt that either popular opinion concerning thebanks or the interest group influence of the German unions was irrelevant.While American politics fragmented capital and labor, German politics broughtthem together in the boardroom.

Codetermination has three important effects on German corporategovernance. First, codetermination makes powerful intermediaries morepolitically palatable in Germany than they have been in America, because theemployees are in the boardroom as a counterweight.

Second, codetermination affects the mechanisms of corporate governanceby, for example, impeding takeovers. In the 1980's the rise of a takeovermarket in America induced popular fears that takeovers would disruptemployment. Antitakeover laws were the result, making takeovers moredifficult. German codetermination has a similar effect; takeovers that woulddisrupt employment are difficult because the shareholders can never capture theentire supervisory board.

Third, codetermination affects and changes corporate governance in thesupervisory board. Obviously, it impedes powerful intermediaries from pushingfor rapid organizational change that would disrupt employment. Moregenerally, bankers know that a powerful supervisory board enhances theauthority of the employees. My understanding is that the bankers have thussought to weaken the supervisory board, the arena where the employees are,while hoping that the managerial board will act as the bankers wish, perhapsafter direct consultation with the shareholding institutions.'39

The rhetoric of banker withdrawal from the boardroom should beconsidered with codetermination in mind. What if banks feared that thegovernance task ahead in Germany is to tighten the belts and salary levels ofhighly-paid German workers? They might plausibly have concluded that the

137. Thomas Raiser, The Theory of Enterprise Law in the Federal Republic of Germany, 36 AM. J.Comp. L. 111, 117 (1988).

138. Alan Hyde, A Theory of Labor Legislation, 38 BuFF. L. REV. 383, 411-12 (1990).139. See supra text accompanying notes 29-30.

1970 [Vol. 102: 1927

Corporate Structure

German political climate would not allow the bankers a visible role in firingemployees or lowering their salaries. Prudence would dictate a lowered profile.

History helps explain the German boardroom and the powerful Germanbanks in another way. After Bismarck unified Germany, he sought to developGerman industry by creating great banks as engines of development. 4 ' Astatist political system facilitated a central bank that provided liquidity to theGerman banks' long-term investments. 4 ' These banks naturally expected tooversee the investments made with funds they lent and thus involvedthemselves in corporate governance. Happenstance helped propel the banks'stock power. The newly unified German state taxed transfers of securities.Stock owners wanted to avoid the taxes; German banks held customers' stockin the banks' name and issued receipts to the retail owners. Then, when onecustomer sold stock to another customer, the banks argued that no taxabletransfer occurred, because a bank was still the record owner. The taxingauthorities agreed. Thereafter, stock owners preferred to deposit stock withbigger banks, which could best match customers' sales and purchases,' 4 thusgiving banks control over the proxy machinery.

One last aspect of German political history is relevant. For the nineteenthand most of the first half of the twentieth century, democratic politics and itsfragmenting tendencies could not affect German intermediaries becauseGermany was not a democratic nation. This is not to say that financial powernecessarily clashes with a democratic political system-democracies haveconcentrated industries and can also have concentrated intermediaries-but itdoes mean that the democratic features fragmenting American financialinstitutions from the nineteenth century to today have had less time to affectGerman intermediaries.

Similarly, the history of German corporate governance is also partly atransmission of government directions through the bank regulators to the largebanks, which implement the directions and get financial protection from thegovernment. (Japanese and German corporate governance and regulation aresomewhat similar in this respect.) This statist structure, while again notnecessarily at odds with democratic politics in general, is at odds with

140. ALEXANDER GERSCHENKRON, ECONOMIC BACKWARDNESS IN HISTORICAL PERSPECTIVE 14-15(1962).

141. Richard Tilly, Banking Institutions in Historical Perspective: Germany, Great Britain and the

United States in the Nineteenth and Early Twentieth Century, 145 ZErrSCHRIFr FOR DIE GESAMTE

STAATSWISSENSCHAFT [J. INSTITuTIONAL & THEORETICAL ECON.] 189-209 (1989). While bank-drivenindustrialization went forward in Germany, the Bank of England would not provide that kind of liquidityto British banks. Id. And due to Andrew Jackson's populist-inspired veto of the rechartering of the SecondBank of the United States, the U.S. in the nineteenth century had no central bank to provide bank liquidity,inducing them to shun long-term, illiquid investments.

142. See generally J. RIESSER, THE GREAT GERMAN BANKS AND THEIR CONCENTRATION, S. Doc.

No. 593, 61st Cong., 2d Sess. 618-620 (1911). Riesser does not say whether the large banks lobbied forthe tax result. The tax also stunted a German securities market. Richard H. Tilly, German Banking, 1850-1914: Development Assistance for the Strong, 15 J. EUROPEAN ECON. HIST. 113, 126-27 (1986).

1993] 1971


American democratic history and its impulse to fragment both concentrationsof private economic power and centralized political power.

B. Japanese Interest Groups

Before World War II, the largest firms in Japan were the zaibatsu, whichresembled the U.S. conglomerates of the 1960's but with family ownership, notpublic ownership, at the top. The zaibatsu were tightly connected to largebanks but the family controlled the banks, which resembled.German universalbanks in their use of long-term loans and equity positions. 4 3 In 1943, theJapanese military disrupted the system by ordering managers at munitionsmanufacturers to follow bureaucratic orders rather than shareholder directivesand by directing munitions firms to choose a main bank to facilitate theauditing of wartime production.' 44 In 1948, orders from the SupremeCommander, Allied Powers (SCAP) destroyed that corporate system withdirectives to break up the zaibatsu, to distribute their stock, and to prohibitbank ownership of big blocks of stock 145 stemming from the "Americanbelief that [democracy] not only required free elections, free speech, and dueprocess .... but also the Glass-Steagall Act."' 146 Little else better shows lawdetermining corporate structure than law imposed by a military dictatorship in1943 or law imposed by an occupying military power in 1948.

SCAP prohibited Japanese banks from owning more than 5% of anotherfirm's stock, foreshadowing the American Bank Holding Company Act of1956. But international politics-communism in China and war in Korea-andfear of Japanese domestic instability changed SCAP's primary goal frompacifying a defeated enemy to building a stable economic ally. So SCAPloosened its grip and decided not to pursue full fragmentation; it watched asJapan fostered close relations between finance and industry. During thefollowing decades, stock relentlessly moved from individuals to banks andinsurers, and cross-ownership bound finance and industry together.

Military directives-a type of law-deeply affected modem Japanesefirms' ownership structure. The wartime Japanese military orders werepractical in origin, but had some anti-capitalist ideology behind them.American occupation orders to force Glass-Steagall rules on the Japanese werenot designed for the efficient operation of financial markets and the

143. WILLIAM M. Tstrrsui, BANKING POLICY IN JAPAN 5, 11 (1988).144. AOKI, supra note 82, at 37.145. See generally J. Mark Ramseyer, Legal Rules in Repeated Deals: Banking in the Shadow of

Defection in Japan, 20 J. LEGAL STUD. 91, 99-100 & n.21 (1991) (comparing Japanese and U.S. bankingregulation).

146. David G. Litt, Jonathan R. Macey, Geoffrey P. Miller & Edward L. Rubin, Politics,Bureaucracies, and Financial Markets: Bank Entry into Commercial Paper Underwriting in the UnitedStates and Japan, 139 U. PA. L. REV. 369, 380 (1990).

1972 [Vol. 102: 1927

Corporate Structure

development of sound corporate structures, but to inculcate democracy,' 47

and the weaknesses now emerging in Japanese intermediaries are thus partlythe delayed result of American democratic ideology.148 Oddly enough,American ideology also accounts for why the Japanese banks were not furtherweakened and were allowed to keep their large size. SCAP initially plannedto dissolve the big banks and end nation-wide branching, but changed its mind.The American occupation bureaucracy thought that dissolving the zaibatsu wasmore important than dissolving the banks; since SCAP's antitrust analysisfound that nine big banks and a fringe of smaller banks provided adequatecompetition, it thought the American model of financial fragmentation wasinapt and thus forced the adoption of only Glass-Steagall but notMcFadden.

49

Although SCAP failed to fragment Japanese finance fully, it did break upthe zaibatsu, limit bank stock ownership to 5%, and segment commercial frominvestment banking. The incompleteness of financial fragmentation meant that,with subtlety, the Japanese could-and did-overcome the American-imposedGlass-Steagall rules-by keeping their banks big, by not adopting Americanpassivity rules, and by channeling postwar credit through the banking system.This they did; with neither interest groups nor populism to block such efforts,the political task was easy. But now, in the 1990's, the banks are doing theheavy lifting in corporate ownership, and the weakening of banks puts thesystem under pressure. Some of these current stresses on banks are delayedreactions to restrictions imposed by U.S. occupation. 15 Although Japaneffectively repealed American efforts at fragmentation by channeling creditthrough banks, such a system cannot continue in the 1990's. Thus, theAmerican-imposed segmentation pressures Japanese financial intermediariestoday in ways it did not in prior decades. If this weakness leads banks to selltheir stock, it could change the structure of ownership and authority in thelarge firm.

To preserve concentrated ownership, Japan could restructure itsintermediaries, but fights among interest groups have stymied a complete

147. See, e.g., id. at 379-80.148. Occupation authorities in Germany imposed McFadden, although German authorities viewed it

as unwise. When the occupation ended, Germany allowed geographically fragmented banks to merge, whichthey did. TsuTsuI, supra note 143, at 55-56; Rolf Ziegler et al., Industry and Banking in the GermanCorporate Network, in NETWORKS OF CORPORATE POWER: A COMPARATIVE ANALYSIS OF TEN COUNTRIES

91, 106 (Frans N. Stokman et al. eds., 1985); Hans A. Adler, The Post-War Reorganization of the GermanBanking System, 63 Q.J. ECON. 322 (1949).

149. TsurSUI, supra note 143, at 41-43 (zaibatsu focus); id. 45-48, 63, 117 (incompletefragmentation); id. at 49-53 (SCAP plans to dissolve large banks); id. at 52, 119 (MacArthur decides banksare secondary).

150. Current Bank of International Settlements capital standards make bank stockholding moredifficult. Japanese savers and corporate borrowers now have access to nonbank alternatives. Interest rates,previously depressed by the government, are reaching market levels. The American Structural ImpedimentsInitiative seeks to fray the equity ties inside the keiretsu.

19731993]


restructuring.' 5 Securities firms see bank underwriting of securities bycommercial banks as dangerous; 52 commercial bankers, not surprisingly, seeno danger.153 The best history of modern financial regulation in Japan seesinterest group infighting, not arguments over the best means of achievingefficiency, as explaining the bureaucracies' slowness in deregulation:

[T]he walls that have divided various types of financial institutionssince World War II still stand, because of the tenacity of entrenchedinterests.... City banks, for example, will be barred from makinglong-term loans in the Euroyen market until the long-term creditbanks and trust banks receive suitable compensation, such asexpanded securities powers. . . . It is the political power of thesevarious groups rather than economic rationale that protectsthem ....

Thus, the newly-weakened Japanese commercial banks may be unable toprevent ownership fragmentation unless they get new authority, which theAmerican-created interest groups so far impede. If full fragmentation doesoccur in Japan, politics-this time, interest group politics-will play some role.

While securities firms want to preserve one part of the current system-segmentation-managers want to preserve another part-cross-holdings, whichprotect them from takeovers and usually give them large friendly shareholders.If banks cannot continue their role in the keiretsu, managers want a substitute,not the unknown. Their representatives argue that "any hard-landing approachwhich restricts ... cross-holding... risk[s] ... demolishing the base of [the]Japanese management systems, and will never be accepted by a nationalconsensus and must be avoided by all means."'155

Political history affects why the American and Japanese banking systemslook so different. A half-century ago, both Japan and America faced bank

151. See ZIELINSKI & HoLLowVAY, supra note 80, at 70 (stating that post-retirement-at 55-jobs keepsome Ministry of Finance officials attuned to interests of banks, others to that of brokers); Elliot Gewirtz& Clark Taber, Fundamental Issues in Japanese Financial System Reform, 7 REV. BANKING & FIN.SERVICES 135, 141 (1991); Litt, Macey, Miller & Rubin, supra note 146, at 404-22 (describing conflictbetween Japanese bankers and brokers concerning banks' underwriting of commercial paper); JamesSterngold, A Japanese-Style "Old Boy" Network, N.Y. TIMES, June 7, 1991, at Dl (discussing efforts ofMinistry of Finance to resolve conflict between bankers and brokers concerning decision to allow banksinto securities business).

152. [Japanese] Securities and Exchange Council, How Basic System Regarding Capital Market Oughtto be Reformed 17-18 (June 19, 1991) (unpublished report, on file with author).

153. [Japanese] Financial System Research Council, On a New Japanese Finance System (June 25,1991) (unpublished report, on file with author).

154. ROSENBLUTH, supra note 79, at 94-95.155. Foundation for Advanced Information and Research, Japan, A Perspective on Japanese Merger

& Acquisition from International Viewpoint 23 (Sept. 1990) (unpublished report, on file with author)(emphasis added); see also RYUTARO KOMIYA, THE JAPANESE ECONOMY: TRADE, INDUSTRY, ANDGOVERNMENT 255-56 (1990); Gerlach, supra note 33, at 117 (describing Japanese executives as opposedto stopping cross-holding). Changes in Japan's pension funds may create a huge emerging pool of capital.RoSENBLUTH, supra note 79, at 75. Whether intermediaries or managers control these funds may determinethe future of Japanese corporate governance.

1974 [Vol. 102: 1927

Corporate Structure

crises. Politics led each nation to react differently. Japan then had thousandsof small banks. Japan also lacked a populist, antibank ethos.'5 6 During the1927 economic crisis, many small banks were badly run and failed, while thelarge ones were fairly well-run and stable. Depositors ran off to thegovernment's savings system, the large banks survived, and the governmentencouraged mergers among the small banks, 157 concentrating banking. Today,Japan no longer has thousands of small banks.

In 1933, America had a similar banking crisis and a different politicalresult. The United States, like Japan, had thousands of small banks. Whenmany of them faced collapse in 1933, they pressed Congress for federaldeposit insurance at the same time Glass-Steagall separation was on theagenda. The interest-group impetus for extensive deposit insurance has beenthe political power of small country banks. They got it enacted,' extended,and beat back attempts to get it under control.159 Without it, Americanbanking would have become more concentrated as many deposits would haverun off from small, weak country banks to larger, often stronger, money-centerbanks. 160 Recognizing this, the large banks supported Glass-Steagallseparation because in 1933 they were not making money in the securitiesbusiness and they hoped this support would deter deposit insurance. 16' Thelarge banks miscalculated, however, and Congress, after passing the Glass-Steagall Act, also passed deposit insurance, which to this day continues to prop

156. T.A. BISSON, ZAIBATSU DISSOLUTION IN JAPAN 29 (1954) (quoting U.S. DEPARTMENT OF STATE,PUB. No. 2628, REPORT OF THE MISSION ON JAPANESE COMBINES, FAR EASTERN SERIES 21 (1946))("Since the [Meiji] restoration .... [t]here has never been any movement in Japan strong enough toproduce a Sherman Act,... a Money Trust Investigation, a Federal Trade Commission, or a Securities andExchange Commission such as developed in the United States .... ).

157. Frances Rosenbluth, Bank Consolidation in Prewar Japan: The Market for Regulation under aNon-Sovereign Diet 10-14, 17-25 (Mar. 1991) (unpublished manuscript, on file with author).

158. Carter H. Golembe, The Deposit Insurance Legislation of 1933: An Examination of itsAntecedents and its Purposes, 76 POL. SCI. Q. 181 (1960); see also Gerald P. O'Driscoll, Jr., DepositInsurance in Theory and Practice, in THE FINANCIAL SERVICES REVOLUTION: POLICY DIRECTIONS FOR THEFUTURE 165 (Catherine England & Thomas Huertas eds., 1988) and sources cited therein; cf. Donald C.Langevoort, Statutory Obsolescence and the Judicial Process: The Revisionist Role of the Courts in FederalBanking Regulation, 85 MICH. L. REV. 672, 695-97 (1987).

159. Kenneth H. Bacon, White House Bill on Bank Law Reform Faces Hurdles as It Goes to HousePanel, VALL ST. J., May 14, 1991, at A24 (reporting Independent Bankers Association's attempts to headoff Treasury effort to limit deposit insurance coverage). Some deposit insurance for the poor and middle-class is socially desirable and could explain a down-sized system, but cannot explain the extensive systemAmerica has. See generally Robert C. Clark, The Soundness of Financial Intermediaries, 86 YALE L.J. 1(1976).

160. Some strength of large money-center banks came from being "too big to fail," meaning that thegovernment-in modem times the Federal Deposit Insurance Corporation or the Federal Reserve-wouldprevent the failure of any really big bank.

161. Langevoort, supra note 158, at 690-91; cf. Jonathan R. Macey, Special Interest GroupsLegislation and the Judicial Function: The Dilemma of Glass-Steagall, 33 EMORY L.J. 1, 2 (1984).

19931 1975


up thousands of small banks. 162 Politics subsidized the small banks, and theUnited States still has thousands of them.

A populist, antibank ethos is historically absent in Japan.'63 LikeGermany, this was partly because until the postwar era Japan was not ademocratic country where such opinions could much influence governmentdecisions. But consider the current turmoil in the Japanese economy. Oneproblem facing Japan is that many firms are in mature product markets andhave substantial free cash flow, which the firms can retain, freeing themselvesfrom the bankers, who first gained influence because they were the necessaryproviders of now unneeded cash. Two tests of the current structure now loom:First, will stockholding banks keep their stock and will they induce portfoliofirms to use the cash well? Second, if the banks succeed but make managersand workers unhappy, say, by ending lifetime employment, a major test of thepolitical theory would arise. In America the disadvantaged groups would tryto shift decisionmaking from the market to the political arena, appealing to thelegislature to re-do the economic result. While in Japan that appeal might fall,politics there is not so different that we should expect disgruntled silence fromthose affected. We shall see.

I make no claim to a satisfactory understanding of the political history ofGermany and Japan and its effects on financial intermediaries and corporatestructures. Although others will have to deepen and correct the history I'vepresented here, even this cursory review shows that political forces haveshaped the German and Japanese corporate structures-forces that differ bothin kind and sometimes in strength from those that shaped American financialintermediaries and American corporate governance. The point is not that theforeign systems are less politically determined than America's. Germancodetermination in the face of revolution is hardly apolitical. The Japanesedecision after 1927 to concentrate their banking system was not apolitical. Thepoint is that different historical politics led to different financial institutionalstructures, and different institutions led to different corporate governancestructures. Today, populism in Germany weakens the power of Germanbankers in the boardroom through formal codetermination and the informaleffects of popular resentment; interest group infighting in Japan and the

162. TsursuI, supra note 143, at 3-4, 11; Juro Teranishi, Financial System and the Industrializationof Japan: 1900-1970, BANCA NAZIONALE DEL LAVORO Q. REV., Sept. 1990, at 309, 329-30; JuroTeranishi, Availability of Safe Assets and the Process of Bank Concentration in Japan, 25 ECON. DEv. &CULTURAL CHANGE, Apr. 1977, at 447,448-49,462,465, 469. Moreover, even earlier, after a bond failurein the nineteenth century, the Ministry of Finance sought to make banking somewhat more concentrated.Gary Saxonhouse, Mechanisms for Technology Transfer in Japanese Economic History, 12 MANAGERIAL& DECISION ECON. 83, 85 (1991) (citing 1 NIHON GINKO HYAKUNEN SHI (A HUNDRED YEAR HISTORY OFTHE BANK OF JAPAN (1982)).

163. BISSON, supra note 156, at 29.

1976 [Vol. 102: 1927

Corporate Structure

American-imposed legacy of partial fragmentation limit the power of Japanesebankers in industry.

V. SHARED AuTHoRrrY: WEAKNESSES AND STRENGTHS

My primary goal in this Article is to show that there are differences incorporate structure, that the German and Japanese structures would be illegalin the United States, and that existing corporate theories, which focus oneconomic factors and do not consider political forces, cannot adequatelyexplain the differences. My goal in this Part is first to find the ways in whichGerman and Japanese corporate governance structures resemble each other,while departing from the American structure, and then to develop hypothesesfor future investigation about the link between structure and performance.

I realize that readers of a competitiveness symposium would prefer that Ioffer a corporate governance silver bullet to cure whatever economic ills afflictAmerican industry. My goal here is more modest. It is not to show that theforeign structures are superior, and hence should be adopted here. Indeed, myessential claim is not that the foreign structures are superior, but that they aredifferent. Moreover, prescriptions for America can be best discerned byexamining American corporate governance. Any governance feature that wouldimprove American corporate governance would make American firms morecompetitive. There is no reason to confine a normative inquiry to features ofthe foreign firms.

Were this Article not part of a competitiveness symposium, I would notinclude a prominent normative section because the current data is far too thinto make concrete recommendations based on the foreign systems. Anyoneadvocating that some feature in a foreign system be adopted here would haveto respond to criticism focusing on the feature's cost, and we lack the datanecessary to evaluate the relative magnitudes of benefits and costs.

I doubt that corporate governance-even if we knew how to achieve theperfect system-is central to economic performance and competitiveness.Macroeconomic policies, competition, industry structure, and the education andmotivation of managers and employees each surely affect competitiveness andproductivity more than does corporate governance. Good performance abroadwill be difficult to trace to corporate governance; and conversely, poorperformance abroad does not mean that foreign governance is worse.I14

164. Thus one might point to America's competitive superiority after World War H, as is pointed outin Roberta Romano, A Cautionary Note on Drawing Lessons from Comparative Corporate Law, 102 YALELJ. 2021 (1993), when the U.S. had, more or less, the same corporate governance system then as it doesnow. So, one might argue, let's ignore corporate governance as a possible American weakness. There weremany reasons for America's relative competitive superiority over Germany and Japan in 1945. It is logicallypossible that America's corporate governance was bad then and bad now, but because governance is atertiary economic feature, it didn't drag down the American economy. Similarly, as I argue below, it ispossible that American corporate governance is, overall, superior to foreign corporate governance now, and

1993] 1977


Moreover, the foreign systems begin with several obvious defects.Japanese cross-ownership and German bank proxy control are afflicted withdeep conflicts of interest, which could deteriorate into mutual managerial self-protection in Japan and opportunism in Germany. Neither foreign system islikely to yield entrepreneurial leadership; committees are not entrepreneurs, andthe foreign institutional allocation of capital seems to do poorly in financingnew entrepreneurial firms. Both foreign systems are associated with heavy statedirection, which has tended not to work well in the United States, and may yetprove to work poorly abroad. My goal therefore is not to show that the foreignsystems have an overall superiority, but to find theoretical benefits that mightoffset the foreign structures' obvious costs.

Any normative inquiry also suffers from the fact that Japanese, andperhaps German, firms may maximize size, not profits. 65 Because one-thirdof the cross-holdings in Japan are held, not by financial institutions, but byindustrial firms, which are as interested in their own sales as in profits fromowning stock, and because employees' representatives make up half of theGerman board and are as interested in jobs as in profits, belt-tightening anddown-sizing would not be easy. If so, the foreign systems may work wellwhen economic determinants impel expansion, but not when expansion is nolonger warranted. The U.S. system may be more adaptable and, when thepluses and minuses are added up, superior. For example, American managersown more stock and America's efforts at pay-for-performance (despite howbadly it is implemented here) may motivate management better than Germanor Japanese approaches. The primary theoretical advantage in the foreignstructures-institutions with very large blocks of stock whose agents interactregularly with management-may disappear in the future if Americanintermediaries continue to concentrate their holdings and become more active.

One last problem afflicts the normative inquiry. How can an increase inbanker power reduce managerial agency costs without increasing bankeragency costs? Empowering financial institutions may improve the performanceof owners' agents in the corporation, but, by expanding the duties of theowners' agents in the financial institution, create new problems. Measuringwhether the gains exceed the losses will be hard. But, perhaps a reduction inagency costs is not the central problem that institutional ownership addresses.Indeed, more than corporate governance-the agency problems of managersarising from fragmented stock ownership-is affected by the structure of largefirm ownership. The problems that institutional stockholding in a flatter

was even then. But that should not stop us from looking at foreign ownership structure to see if it offersany lessons for America. I see no reason why we should not try to take the incremental gains and leavebehind the losses, if we can.

165. Alan S. Blinder, Profit Maximization and International Competition, in 5 FIN. & INT'L ECON.:AMEX BANK REv. PRIZE ESSAYS 37 (Richard O'Brien ed., 1991).

1978 [Vol. 102: 1927

Corporate Structure

hierarchy might mitigate fall into three categories: problems in management;problems in financial markets; and problems in organizing large scale industry.

A. Problems with Management

Contrary to conventional wisdom, bank ownership abroad did not shiftcontrol from managers to banks, but led to power-sharing. Day-to-day banker-shareholders do not run the firm; they intervene in crises and may holdmanagers accountable in the interim. Managers presumably try to avoid thosecrises and the threats these blocks bring to their own autonomy. Second,several intermediaries usually hold big blocks of stock in large firms-no oneintermediary dominates. Third, the intermediaries are highly leveragedinstitutions that cannot afford to make a serious error. Fourth, the stockholderis personified, changing the way managers view owners. Fifth, shared authorityallows more people from different organizations into decisions at the top,reducing inbred decisionmaking.

1. Increasing Accountability

Although reducing agency costs may not be the key advantage, if there isone, to the foreign systems, increased accountability should not be overlooked.American managers have historically been less accountable to shareholdersthan their foreign counterparts, 166 despite big gaps in foreign accountability.While the trend may be toward less accountability abroad and moreaccountability here, differences still exist.

Thus, a first hypothesis is that the German and Japanese corporatestructure motivates the board to be more conscientious. The German andJapanese CEO must meet with the representatives of big stockholders who donot always owe their positions to the CEO, but owe them to their homeinstitutions, and who do not depend financially or socially on the CEO, adependence that afflicts American directors. 167 In Germany, managers, whohandle day-to-day matters, interact with institutional stockholders whenreporting to the supervisory board of directors in regular, formal meetings.German bank directors, unlike some American directors, are not appointed tothe board because they have a reputation for passivity, but frequently becausetheir institutions vote big blocks of stock. And in Japan, managers dominatethe internal board that handles day-to-day matters but interact with bigstockholders in regular, informal Presidents' Council meetings. Key members

166. Cf. Bernard S. Black, The Value of Institutional Investor Monitoring: The Empirical Evidence,39 UCLA L. REV. 895, 898-916 (1992) (discussing gaps in American accountability).

167. JAY LORSCH & ELIZABETH MACIVER, PAWNS OR POTENTATES: THE REALITY OF AMERICA'SCORPORATE BOARDS (1989); Victor Brudney, The Independent Director-Heavenly City or PotemkinVillage?, 95 HARV. L. REv. 597, 598 (1982).

19931 1979


of the supervisory board and the informal Presidents' Council owe theirallegiance to their own institutions and thus have reason not to follow the CEOblindly. Moreover, because the CEO cannot join the German supervisory boardand faces several big stockholders in the Japanese Presidents' Council, theCEO is not able to co-opt the external monitors completely. Thus, in bothforeign nations, managers dominate an internal board, but not the externalsupervisory board or the Presidents' Council. In America, in contrast, corporategovernance involves only a single board, which the CEO has historicallydominated.

Although neither foreign meeting structure facilitates a detailed review ofsenior managers-German supervisory boards meet a few times per year,usually without a searching agenda, and Japanese monthly Presidents' Councilmeetings involve so many firms that no one firm will usually be the focus ofattention-these meeting structures are fit for crisis management and big-picture, ongoing review. Hence, even though foreign CEO's are not subject today-to-day direction from institutional investors, as is sometimes thought, theyare weakly accountable to institutions whose continuing involvement makesthem knowledgeable about the CEO's business. Managers get their chance, andthe intermediaries intervene primarily when managers fail. Certainly data onexecutive turnover shows that German and Japanese ownership does not deterremoval of executives when performance slackens;'68 involuntaryresignations increase when the Japanese economy weakens. 169 This beliesassertions that the big stockholders insulate managers from removal. Thequestion remains whether these removals are fast enough or occur in the rightcircumstances, and whether the removals actually improve corporateperformance. Large stockholders' ongoing involvement means they do not needtime to move up from the bottom of a learning curve and can intervene quicklywhen crises hit. Finally, the potential for institutional interruption in crisisshould improve firm performance as managers try to avoid those crises. 7 '

2. Monitoring by Multiple Intermediaries

German and Japanese firms usually have several intermediaries holding bigblocks' of stock. Multiple intermediaries can (1) deter opportunism bymonitoring one another; (2) impel action in a way that a single blockholdermight not; and (3) facilitate power sharing, not domination. Rarely does a

168. See Kaplan & Minton, supra note 36; Steven N. Kaplan, Top Executives, Turnover and FirmPerformance in Germany (Feb. 1993) (unpublished manuscript, on file with author).

169. See supra notes 36-37.170. These elements of shared authority comport with agency theory, which suggests that organizations

can separate some management and control functions internally by (1) using a hierarchy in which asubordinate initiates and a superior ratifies and monitors; (2) using a board that seriously reviews topmanagers; and (3) mutual monitoring across decisionmakers. Jensen & Smith, supra note 8, at 101.

[Vol. 102: 19271980

Corporate Structure

single dominant bank own a control block. Thus, the foreign structures do notreplace an American-style commanding CEO with a foreign-style commandingbank officer.

a. Multiple Blocks Checking Each Other

A single bank in control of a firm could charge an above-market interestrate, compel the firm to do fee-generating deals, divert value from the firm toitself, or pursue its own empire-building agenda.' 7 ' First, because eachsystem tends to have multiple intermediaries-Deutsche Bank's dominating28% block of Daimler-Benz is exceptional-the risk of bank opportunism isdiminished (although hardly eliminated). German firms usually borrow fromseveral banks and German loan rates are said not to include a premium forbanker control; 172 Japanese firms occasionally end their relationship with themain bank and seek another.173 To form a coalition, the bank must appeal tothe others. That appeal may be opportunistic, but the multiple parties involvedgive managers the chance to break an opportunistic coalition. Since themanagers control some stock in the financial institution, they have a base uponwhich to build a counter-coalition to oppose the opportunist. A true managerialcrisis inside the firm, however, should unite the stockholders-all want to endthe crisis-but the opportunist may have trouble acting alone. 74

b. Multiple Blocks Impelling Action

Why don't American financial intermediaries-some of which can ownsome stock-go to the law's limits in stockholding? Isolated blocks, even bigones with 5% of the portfolio firm, are precarious because management mayisolate, outmaneuver, and destroy them. 75 In America, a single 5% blockdoes emerge here and there, but managers still have the upper hand in timesof conflict. Even a half-dozen institutions with 5% blocks would have tocoordinate their activities, which U.S. securities regulation historically deterred.

171. See Stuart Rosenstein & David Rush, The Stock Performance of Corporations That Are PartiallyOwned by Other Corporations, 13 J. FIN. REs. 39 (1990).

172. John Cable, Capital Market Information and Industrial Performance: The Role of West GermanBanks, 95 ECON. J. 118, 118-19 (1985); see also Eckstein, supra note 126, at 478.

173. Ramseyer, supra note 145; Toshiaki Tachibanaki & Atsuhiro Taki, Shareholding and LendingActivity of Financial Institutions in Japan, 9 BANK JAPAN MONETARY & ECON. STuD. 23, 24, 31 (1991)(reporting that one in four firms switched main banks in a four-year period).

174. Even this is not a complete explanation. Institutions might form coalitions. Banker A may agreewith Banker B that A will exploit company X while B exploits company Y. Nonetheless, such coalitions,like cartels, are often difficult to construct and to maintain. Moreover, bankers should fear that visibleopportunism will provoke political retaliation, particularly in Germany. Politics constrains.

175. Large stock positions in big American firms are rare. One example was H. Ross Perot's 6% stakein GM. Perot gained access to the GM Boardroom, but when the GM managers discovered that he was afeisty critic, they threw him out. Perot was not able to form an alliance with other large stockholdersbecause there were no others. See Table I.

1993] 1981


Thus, some "unleashed" American institutions should rationally refuse to holdthe maximum permitted by American law because they know that in times ofcrisis they will need a critical mass of other large blockholders, a critical massthey lack.

The ownership structures in both Germany and Japan have weaknesses inimpelling action. The German banks' big voting blocks usually come frombrokerage stock, for which the German banks are not likely to make heroicefforts at monitoring. 76 Each Japanese bank owns no more than 5% of anindustrial firm's stock, a position that is frequently too weak to wrest controlfrom managers in the United States. A German bank's typical directly-owned,nonbrokerage position, while larger, is usually too small standing alone to givethe bank control; with brokerage stock or alliances with other stockholders, thebank can obtain control.

Multiple blocks may induce intermediaries to act in a way that a single

isolated blockholder might not. The multiple blocks may solve the weaknessproblem, without creating a dominating block that recentralizes authority in adominbting intermediary. The roles of German brokerage stock and of multipleJapanese 5% blocks may not be to share expense with the active intermediary,but to give voting power. Directly-owned stock gives the institutional ownera financial incentive to act; ancillary, backup stock bulks up voting power toprovide the means.'77

c. Multiple Blocks and Power Sharing

If foreign intermediaries are large Berle-Means firms (actually many arenot),t7 8 they could have too many scattered shareholders to give their agentsthe proper incentives. Why should reducing managerial agency costs at thefirm not just displace the problems up one level into financial intermediaries,making bankers the source of a new agency problem?

This question may make the normative inquiry a dead end. But powersharing differs, at least in form, from completely shifting authority frommanagers to bankers. 79

176. High registration fees drive individuals' stock into banks, whose custodial fees are not tied toperformance. See BAUMS, supra note 16, at 25-27. The banks could be paid via conflicts of interest thatallow them to divert resources from the firms to themselves, presumably via favorable loan terms or highfees for deals, but the current evidence says this does not in fact happen. Cable, supra note 172, at 119-24.

177. If political pressure on German banks leads them to reduce their amount of directly-held stock,see supra text accompanying notes 130-133, while retaining their brokerage stock, they may face reducedincentives to act responsibly.

178. The Japanese banks are primarily cross-owned by other Japanese financial institutions andindustrial companies; in Germany, one of the three big banks is also partly cross-owned by a largeinsurance company and industrial firm.

179. Shared authority affects monitoring, a fact that is this Subsection's focus. It can also affect thequality of decisions, the focus below. See infra text accompanying notes 185-188.

1982 [Vol. 102: 1927

Corporate Structure

This shared authority lies between market and hierarchy, linking financeand industry in a complex way. Control is not the right word to describe theserelationships. Interlock, or Escher-like overlap, might be more apt.' AnEscher-like hand reaches out from the bank to control the firm, but then anEscher-like hand reaches back from the firm to control the bank.

To see this, trace the historical authority of the Japanese banks. In theirheyday, through their control over credit, they could control firms, review theirbudgets, and monitor their managers. They owned stock in the monitoredfirms. rhe assets of the monitored firms included the stock of the monitoringbanks. The banks could influence firms through their stockholdings, and thefirms could influence the banks through their own stockholdings. The banksthus were not (and are not now) precisely Berle-Means firms.

Escher-like overlap also describes corporate governance in Germany,although less clearly. Bankers have seats on industrial firms' supervisoryboards and managers from important industrial firms are members of thebanks' boards and on the banks' informal national advisory boards.'Moreover, managers direct employees, some of whom sit on the supervisoryboard and direct managers. Corporate governance structures in both nationsresemble less a hierarchical pyramid-the American picture-than an Escher-like staircase-while always walking downstairs, we wind up on top of thestaircase from which we started.

This interlocking, shared power model comports with the Japanesestructure in two ways. No single bank dominates a firm's stock ownership;only a group can dominate. Moreover, groups of firms cross-own someinfluential blocks of stock in the banks. This Escher-like interlock modelcomports with the German structure in one way-only a coalition cancompletely dominate the firm. Two of the three large German banks are,however, themselves Berle-Means firms with scattered shareholders, aconsideration we analyze next.

3. Monitoring the Banks

Ignore cross-ownership of the banks for now-as we must for two of thethree large German banks and as we can for the Japanese banks-byconsidering institutional ownership's effectiveness in another setting. How canthe banks monitor when they themselves are presumably afflicted with agencyproblems? As we have said, this may make this part of the normative inquirya dead end. But consider the banks as embedded in a simple tri-level

180. M.C. Escher was a Dutch graphics artist with whose work most readers will be familiar. Escher'sdrawings show, for example, realistic renditions of staircases that descend to the top of their own ascentand walls that abut ceiling and floor at the same joining point.

181. Kilbler, supra note 133, at 109. The new ownership interlock between Dresdner and Allianzstrengthens the overlap.

1993] 1983


hierarchical monitoring model with industry at the bottom, financial institutionsin the middle, and government regulators at the top. Although theintermediaries monitor industry, they are themselves free from day-to-daycontrol. So, who monitors the intermediaries? No one, day-to-day. But thegovernment monitors leveraged intermediaries and, when it gets a signal offinancial weakness from the banks, it begins inquiries that weaken bankmanagers' autonomy.18 2

In this model then, banks' weakness-high leverage with thinequity-becomes a strength, if properly regulated. Leveraging and regulationcause the senior bank managers to get good returns and avoid financialdisasters at portfolio firms in order to avoid a thinning of the bank's ownequity, which would trigger regulatory intervention. Bank ownership increasesthe accountability of managers, but without an equal and offsetting increase inagency problems in the stock-owning bank because the leveraged bank hasfinancial and regulatory reasons to avoid a severe decline in its stockportfolio.,, 3

4. Personifying Shareholders

American managers owe fiduciary duties to an abstraction, a faceless stockmarket. In Germany and Japan, on the other hand, this abstraction ispersonified through regular interactions with bank directors (on the supervisoryboard) or the bank presidents (at the council meeting). The derelict foreignCEO risks not just betraying abstract duties to an anonymous market, but alsobetraying people from a stockholding institution. Loyalty to real people maymotivate better than legally mandated loyalty to an abstraction.

German and Japanese managers, due to constant interaction with peoplefrom stockholding institutions, should feel greater peer pressure, guilt aboutshirking, more camaraderie, and more empathy among coworkers, all of whichcould improve performance.' 4 CEO's may be more willing to hurt ananonymous stockholder, whose needs are not vivid and present, than they canhurt a cohort. Isolated American managers, on the other hand, are likely to see

182. Cf. Douglas W. Diamond, Financial Intermediation and Delegated Monitoring, 51 REV. ECON.STUD. 393 (1984); Michael C. Jensen, Agency Costs of Free Cash Flow, Corporate Finance, andTakeovers, 76 AM. ECON. REV. (PAPERS & PROCEEDINGS) 323 (1986).

183. This may seem like a problem with institutions and hence could belong to the next Section, whichdiscusses problems with financial markets. But it is best thought of as a problem of managementaccountability that bank ownership theoretically can help, without creating equal and offsetting problemsin the upstairs institution.

184. See Eugene Kandel & Edward P. Lazear, Peer Pressure and Partnerships, 100 J. POL. ECON. 801(1992); cf. Strong v. Repide, 213 U.S. 419 (1909) (though insider stockholder can mislead the market, aface-to-face seller cannot). The risk of course is that personification leads to a failure to hold managersaccountable for their mistakes.

Sociologists found few Americans soldiers fought "for freedom," "for the American way," or "againstaggression"; motivation came not from these abstract ideals, but from loyalty to peers in the platoon andfear of embarrassment in front of them. JOHN KEEGAN, THE FACE OF BATmE 50-53, 72-73 (1976).

1984 [Vol. 102: 1927

Corporate Structure

shareholders, not as friendly institutions expecting a profit, but as anonymoustraders with little loyalty to the firm. American CEO's may see even thenewly-emergent institutional investors having stable but small blocks as toodistant to merit loyalty and respect-a loyalty and respect that could arise fromregular interaction in a governance or quasi-governance setting, such as in thePresidents' Council or in interaction with the supervisory board.

5. Improving Decisions

Shared authority may increase managerial accountability, but that is not itsonly potential benefit. Organizational theory suggests another reason why thispartial vertical integration may be useful. Complexity reduces an individual'sability both to comprehend all that is necessary and to avoid bias fromoutmoded experience. In a modern, complex economic system, information istoo dispersed; no individual or staff can have all of the information needed fordecisionmaking. Networks can do better, the theory goes, than any oneindividual or hierarchy. Moreover, since a decisionmaker's biases are ofteninvisible to the decisionmaker, a single individual will do worse than anetwork of decisionmaking, which reduces error, 15 similar to the way a goodlaw firm with many high quality people in overlapping fields can cooperate,converse, and get the job done better than a lawyer of equal quality workingalone. Unlike the American conglomerate, foreign structures do not replace theCEO with a new centralized decisionmaking authority. Cross-holding allowsmore interested parties to participate in the decisionmaking at the top. In somesettings, wider participation may help; neither the individual CEO nor theheadquarters at the top of firms in technologically complex and fast movingindustries may be able to assimilate all the important information needed forcritical decisions.

The problems of misinformation and bias can be generalized. Markettransactions have costs; command and hierarchy replace market costs withorganizational costs. Relative cost determines organizational structure; firmsarise when a hierarchical structure provides a more efficient means oforganizing than transactions in the marketplace.186 The foreign systemspermit hybrids that could perform better than either pure type.

Changing underlying economic conditions could make hybrids moreimportant than they once were. Increasing complexity of information is one.New theories of firm organization describe American-style command-and-control hierarchy as poorly adapted to today's continual information shifts and

185. See Raaj K. Sah, Fallibility in Human Organizations and Political Systems, 5 J. ECON. PERSP.67, 69-71 (1991).

186. WILLIAMSON, supra note 4, at 82; R.H. Coase, The Nature of the Finn, 4 ECONOMICA (n.s.) 386(1937).

1993] 1985


technological changes. 87 Although the pyramid of decisionmaking workedwell when mass production of slowly changing products was typical, thedifficulties of producing today's complex products in rapidly changing marketsmay make a flatter authority structure at the top better.

Some cross-owned firms in Japanese keiretsu are vertically related andwould be part of a single large vertical firm in the United States. By involvingmore individuals in decisionmaking at the top, the flatter authority structurepresent in German and Japanese firms may serve better to motivate managers,particularly those who would be farther from the top in the American verticallyintegrated firm. Indeed, "[r]esearch suggests that hierarchical design dampensemployee motivation because individuals are.., more committed when theyhave participated in a decision, and much less enthusiastic when ... orderedby superiors to undertake a particular task."'' 8

1

The incentives, sense of responsibility, status, and hence motivation ofsenior managers at semi-independent keiretsu firms should differ from thoseof managers who run American-style divisions or wholly-owned subsidiaries.Keiretsu CEO's should be able to call on stockholder-financiers directly, whiledivision managers and subsidiary heads must first speak with their firm bosses.

Cross-ownership may also help motivate lower-level employees bysupporting the implicit lifetime employment contract that is only nowweakening in the core keiretsu firms. Although stockholders without any otherrelationship to a firm will sometimes renege on an implicit agreement,stockholder-suppliers might honor the agreement because the gains they reapas reneging stockholders would be reduced by the losses they incur as affectedsuppliers. Suppliers may fear demoralization of their own employees, whowould have to deal with some of the victims, employees of the customer-firmwho no longer benefit from lifetime employment. Also, as stockholdersinvolved with multiple relationships-suppliers and stockholders, creditors andstockholders-they may avoid high variance strategies that often benefitstockholders but hurt creditors, suppliers, customers, and employees. To protectthemselves as customers, suppliers, or creditors, cross-holding stockholdersmay protect employees even if, in so doing, they decline to take risks thatwould maximize the value of their stock.

187. Peter F. Drucker, The Coming of the New Organization, HARV. BUS. REV., Jan.-Feb. 1988, at 45,53.

188. Walter W. Powell, Neither Market Nor Hierarchy: Network Forms of Organization, 12 RES.ORGANIZATIONAL BEHAV. 295, 319 (1990).

1986 [Vol. 102: 1927

Corporate Structure

B. Problems with Financial Markets: Institutions as the Problem

1. Enhancing the Information Flow

The ownership structure in Germany and Japan should improve the flowof information from inside the firm to large shareholders, thus helping to deterthe short term propensities often seen in the stock market.'89 To do this inthe United States, managers would have to reveal complex, proprietaryinformation to the stock market, and derivatively to competitors.' 90

Moreover, evaluations of technical data from inside the firm may requireregular, private interaction between large stockholders and managers, which theforeign structures provide but the American structure does not. 9'

2. Improving Loan Markets

Bankers may buy stock in their debtors to create an additional mechanismfor protecting their loans or providing a secondary information channel tosupport their loans. 92 Bankers may also buy stock to limit their ownopportunism-a threat not to rollover short-term loans unless the debtor grantsconcessions is less credible when carrying out the threat would reduce thevalue of the stock the banks hold. Finally, because bankers tend to beconcerned with loan repayment and not with maximizing firm value, they andtheir debtors may both prefer that they buy stock to mitigate theirconservatism. 93

Lastly, when creditors also have large stock positions, workouts andrecapitalizations to avoid a complicated and costly bankruptcy proceeding areprobably easier.

189. I first argued this possibility in Roe, Political Theory, supra note 1, at 55-56.190. Bruce Atwater, the savvy CEO of General Mills, admits that he tries to learn of his competitors'

activities from analysts. Bruce Atwater, The Governance System Is Sound, DIRECTORS & BOARDS, Spring1991, at 17, 19. Presumably CEO's severely limit the amount of proprietary information they release toanalysts.

191. Takeo Hoshi et al., Corporate Structure, Liquidity, and Investment: Evidence from JapanesePanel Data, in ASYMMETRIC INFORMATION, CORPORATE FINANCE AND INVESTMENT 105 (R. Glenn

Hubbard ed., 1990); Stephen D. Prowse, Institutional Investment Patterns and Corporate FinancialBehavior in the United States and Japan, 27 J. FIN. ECON. 43 (1990).

192. Obviously, the influence from stock may yield the mutual self-protection that weakens the

German and Japanese systems: The stockholding bank may induce the portfolio firm to borrow on termsunfavorable to it.

193. Insider trading helps to explain why banks buy stock, or at least what happens once they ownit, particularly in Germany, where, due to a lack of insider trading laws, bank officers can and do trade on

inside information from firms. While insider trading surely explains some stockholding, it cannot explainall of it, because of the Japanese banks' stability in ownership (insiders have to trade to make profits), thetiming of their acquisitions (they all bought in more or less at the same time, to stabilize firms in the faceof American strength and takeover possibilities), and the uniformity of 18-20% bank ownership in the topcompanies (every firm cannot be a better than average deal).

198719931


C. Problems in Large-Scale Industry: Organization

Credible commitment analysis helps to explain the role of cross-ownershipin Japan, where one-third of the cross-ownership of stock involves suppliersand their customers. 94 A supplier wants protection before investing in newmachinery to make a good that only a specific customer can use. What willstop the customer from extorting the supplier later on, after the supplier has themachine? Although a detailed contract between the supplier and the specificcustomer may protect the supplier, many of the ways that the supplier can beexploited are unforeseeable. Vague promises from a customer to act in goodfaith are often not enough. Multiple cross-holdings of stock, however, maymitigate opportunism. If the customer tries to mulct the supplier after thesupplier has committed itself, a coalition of stockholders could intervene tostop the opportunism. 95

A bank's investment in a long-term loan' 96 resembles a supplier'sinvestment in complex machinery for a specific customer. The borrower mayrenege on the loan by increasing the level of risk or, as noted above, the lendermay renege by refusing to rollover the loan at a critical juncture. In the UnitedStates, detailed loan agreements mitigate these problems. Perhaps in Japancross-ownership limits borrower-lender opportunism and thereby facilitatestrade in capital.

197

In both supplier-customer relationships and lender-borrower relationships,the relational commitment is made credible by the plausibility that a coalitionwill form to control the opportunist. Cross-ownership among related customersand suppliers-whether of machinery or of capital-resembles a hostageexchange in which each side gives the other a hostage whose loss will becostly to the offering party but whose value to the side holding the hostage islittle. By acting opportunistically, a supplier risks not only the directintervention of the others but, because the supplier is bound in a web of cross-holdings and relational investments, the supplier will devalue these otherimportant relationships. If a supplier acts opportunistically, the customer cansell the supplier's stock in the market and encourage the other firms in thecross-holding web to sell their stock in the opportunist. Such sales would hurtmanagers at the opportunistic supplier, who would be "naked" and at the risk

194. Yusaku Futatsugi, What Share Cross-Holdings Mean for Corporate Management, ECON. EYE,Spring 1990, at 17, 18. Industrial ownership of stock is also significant in Germany.

195. See Gilson & Roe, supra note 100.196. Some of these loans may, in form, be short-term loans, particularly in Japan. When these loans

are rolledtover regularly, they effectively become long-term loans.197. In fact, this scenario is quite complicated. If fiduciary duties are weaker and more difficult to

enforce in Japan than in America, the Japanese system may require cross-ownership to function. Thus, theJapanese may bear the costs of cross-ownership-whatever they may be-because they lack othermechanisms for enforcing contracts and fiduciary duties.

1988 [Vol. 102: 1927

Corporate Structure

of takeover. The selling customer recovers the market price of the stock; theopportunist is, however, made to operate in a riskier environment. 98

Relational contracting problems may overlap with corporate governanceproblems. Multiple relationships, such as suppliers and customers owning eachother's stock and trading goods, and lenders and borrowers owning eachother's stock and trading loans, double up (1) the sources of information; (2)the incentives to participate in the governance of the firm, since a failinginvestment is also a failing customer; and (3) the means of intervention,allowing for both exit and voice.

The point of the inquiry in this Part is not to prove foreign superiority, but(1) to suggest where to look for their strengths; and (2) to see that there is aprima facie case that their debilities-perhaps they will be slow-moving whendown-sizing is warranted, are likely to overexpand when expansion isundesirable, or may fail to check errant managers due to mutual protection, notmutual checking-have some offsetting strengths that have so far enabled theforeign systems to survive.

VI. IMPLICATIONS FOR AMERICAN STRUCTURE

Corporate structures in Germany and Japan have several commonelements, including shared authority at the top, large financial intermediariesthat hold concentrated blocks of stock, interaction of bankers and managers instructured settings, multiple intermediaries that split the vote, and personifiedshareholders. These common elements differ from American structure. Andwhile the structures face severe risks of stagnation, they potentially have somehypothetical offsetting advantages in improving management, financial markets,and industrial organization.

But hypotheses are not proof. We lack a powerful theory, backed by cleardata, showing how big blocks of stock might improve corporate performance.Thus, the point of comparing corporate structure in Germany and Japan on theone hand and the United States on the other is not to show that the foreignstructuie is superior. Instead, the central point of this Article is to show thatthe American governance structure is not inevitable, that alternatives areplausible, and that a flatter authority structure does not disable foreign firms.But that is not much reason to take the next step and recommend massive legaland economic change. A prima facie possibility is not a mandate for massivechange. Even if some foreign firms are helped, it is doubtful that all are.

198. The sale can be seen not only as a "killing" of the hostage by destroying the mutually protectiverelationship, but also as a direct enforcement mechanism. The stockholders could oust the incumbentmanagement directly by using their stock. In the sale alternative, they oust the opportunistic firm from thenetwork, and leave the opportunistic managers to the perils of the market, which might then oust themanagers.

1993] 1989


Although we do not yet want a legal regime that requires or even encouragesAmerican replication of the foreign corporate structures, we should want alegal regime that permits them. Right now, we do not have one that permitsthem.

A. Germany and Japan as Blueprint?

Yet rules that merely permit alternative structures might be pointless. Forreasons that are not yet obvious, the international pattern is that a nation'slarge firms either all have concentrated institutional voice, or none do. TheJapanese pattern in Table Ill does not reveal diversity, and nearly all Americanlarge firms have ownership structures like GM's in Table I1.199 Economicdeduction suggests that firms that would benefit from concentrated ownershipby financial intermediaries would have it while firms that would not benefitfrom such ownership would have scattered ownership. 200 But theinternational results-either all one way or all the other-suggest that a slighttilt in the power of intermediaries in either direction leads to concentratedvoting or its absence. A nation might not be able to reap the gains from greaterinstitutional voice where it works well without suffering losses where itdetracts from performance-one more reason why this dimension for corporategovernance reform may be a dead end.

Institutional influence here will not come primarily from banks, as it doesabroad. As such, while we can make a case for tearing down the highestbarriers to institutional involvement, I doubt that banking laws will be thecenter of such reform. Moreover, the political economy of American bankingmakes 'reform hard. There are ample reasons in favor of interstate branching,and remarkably few against it, but proposals to repeal the McFadden Act thusfar have failed, largely because small-town, independent bankers form apowerful lobby. Second, because they have a highly leveraged structure, evenunregulated banks will not become major direct equity players. 20 ' To do sothey would need at least modest direct holdings and larger holdings for others'benefit, like the holdings of the German universal banks. But, even ifAmerican intermediaries wanted such holdings-and the riskiness of havingthem makes such an inclination unlikely-such a new complex intermediarywould require regulation that the American system cannot yet provide, andhence American regulators must prohibit it. Third, if reform promises tounleash bankers-an unlikely prospect-savvy managers, who were

199. See Table XIn.200. See Demsetz & Lehn, supra note 22 (describing sorting trend among American firms). IfDemsetz

and Lehn are right, sorting is plausible even if we loosen up institutional constraints.201. Herwig Langohr & Anthony M. Santomero, The Extent of Equity Investment by European Banks,

42 J. MONEY, CREDIT & BANKING 243 (1985). Yet the leveraging, as we saw in Part IV, may propelsuccessful monitoring.

1990 [Vol. 102: 1927

Corporate Structure

instrumental in defeating takeovers that challenged managerial autonomy,"'might weigh in to pui bankers back on their leash. Fourth, fractional reservebanking is a declining industry-not the place to look to for future growth.' 3

Fifth, the path-dependent development of American banks makes them ill-equipped for the tasks of governance. German and Japanese banks "grew up"in the boardrooms (or Presidents' Councils) of their nations' industry;American bankers did not.

In deciding questions of bank size and solvency, we should not-and willnot-give much weight to corporate governance issues. Bank solvency is tooimportant. America's extensive deposit insurance (and the too-big-to-faildoctrine) makes banks and stock a volatile mixture. The biggest risk ofunleashing American banks is less that they will disable industry than that theywill disable themselves and pass these risks onto the public through depositinsurance. Reforming deposit insurance, however, has thus far proven to benearly intractable-again, small banks have too much power. The Bradyreforms were killed shortly after they were announced, 2°4 and Congress hasdone little to limit deposit insurance.2 5 Congress' new effort to control theimpulse to pay uninsured depositors at banks that are "too big to fail" isuntested. Since we cannot unleash banks at the same time that we haveunleashed deposit insurance, a plausible beginning might be to allow bankstock ownership, but only with very high capital requirements, which maymake banks' participation in corporate governance merely aspirational for theforeseeable future.

The historical path of America's weak banking system may mean thatclose relations between America's largest banks and industrial firms areunlikely to develop. Evolution from today might take another path, however.The Federal Reserve could and should revoke its passivity promulgations;

202. Mark J. Roe, Takeover Politics, in THE DEAL DECADE: WHAT TAKEOVERS AND LEVERAGEDBuYOUTS MEAN FOR CORPORATE GOVERNANCE 321 (Margaret M. Blair ed., 1993).

203. Steven Lipin, Bank Industry Seen Shrinking in 1990s, Survey of Executives, Regulators Finds,WALL ST. J., Oct. 2, 1991, at A14. Some of the problems facing the foreign banks partly arise from thepressure on fractional reserve banking all over the world.

204. Ann Devroy & Kathleen Day, Deposit Fee Draws Wave of Protest, VASH. POST, Jan. 26, 1989,at Al.

205. The FDIC Improvement Act, which requires risk-based premiums by 1995, limits payments abovethe explicit insurance ceiling, and encourages regulatory intervention prior to insolvency, may seem toreduce the impact of congressional failure to deal with the core problem of a high ($100,000) insuredamount. FDIC Improvement Act of 1991, Pub. L. 102-242, 105 Stat. 2236 (1991) (codified in scatteredsections of 12 U.S.C.). But we have yet to see how successful these reforms will be. Moreover, we do notyet have reason to believe that regulators will be able to act swiftly and effectively. America's proceduralsystem could hinder even highly capable regulators. And America's political system allows affected banksto appeal to politicians to stop the regulators. See, e.g., Perspective-S&L Scandal, CHI. TRIB., Dec. 9,1988, at 26 (Speaker of the House Wright collected campaign contributions from the S&L and real estatelobbies, and delayed closing down insolvent S&L's). Furthermore, new efforts to control the impulse topay uninsured depositors of banks "too big to fail" are untested. FDICIA § 141 (a)(1)(C) (1991), 12 U.S.C.§ 1823(c)(4)(E) (1988). Hence, I would rather see the system work effectively before we undertake seriousdeposit insurance risks, than to assume that shutdown reform will work and therefore the risks aremanageable.

1993] 1991


Congress could and should remove the 5% lid on holding company stockownership and should instead tie stock ownership to holding company capital.Although few money-center banks are in a position to take advantage of thesechanges, many regional banks are in good shape206 and may take advantageof the new opportunities to become more involved with the governance ofmedium-sized companies, and some of these firms will grow to becomeAmerica's largest companies.

Because American banks do not, and will not for the foreseeable future,directly hold big blocks of stock in America's largest firms, the focus ofAmerican reform will be pensions, mutual funds, insurers, and securitiesmarkets, not banks. Thus, the lessons from abroad-where bankers areprimary-are at best general and superficial. 207 Blueprints for Americanreform will come from studying America's pensions, mutual funds, insurers,boardrooms, and securities markets, not from studying the German or Japanesebank-centered system.

B. Path Dependence and the Futility of Legal Change?

Nor is it clear that legal change will lead to structural change. Pathdependence makes it hard to sort out what the "natural" result is: Americasuppressed powerful intermediaries, but Germany and Japan not only permittedthem, but sometimes encouraged them. During World War II, the JapaneseWar Ministry required munitions manufacturers to choose a main bank; afterGerman unification in the nineteenth century, the German state encouragedbank power and, perhaps accidentally, pushed custodial stock into the handsof the large banks. While banker stock ownership may well have beenancillary to the encouragement of bank power in both nations, it happened.And conceivably even were regulation in Germany, Japan, and the UnitedStates now made to be identical, the path-dependent prevailing structures ineach-bank weakness here, and bank power abroad-would continue.

First, the social gains from involvement are probably not large. Othermechanisms substitute, albeit imperfectly, for large shareholders in theboardroom. Second, due to free rider effects, the private gains are even lessthan the social gains. The institution cannot capture all gains, and thetransaction costs of transformation could exceed the private gains to a 5%stockholder. Third, complex institutions are shaped by their history. Forexample, American large life insurers were barred from buying any commonstock for most of this century. Yet, even with new-found permissions after

206. See, e.g., Saul Hansell, Despite Sharply Different Styles, NCNB, First Union and Wachovia HaveDriven Each Other Into the Top Tier of U.S. Banking, INST. INVESTOR, Nov. 1991, at 101 (discussingregional branching of three of America's largest banks).

207. Some of these other institutions, such as pensions, are tied to banks, which sometimes managepensions.

1992 [Vol. 102: 1927

Corporate Structure

1984, they lack the big investments that insurers have in Germany, Japan, andGreat Britain.

Britain is another example of path dependence reducing the impact of legalchange. It has had a history of partial restrictions: (1) British banking lawprecluded banks from making long-term investments in the nineteenth centuryand early twentieth century; 208 (2) British banks could own stock in thetwentieth century, but needed the Bank of England's permission, which until1980 the Bank would not give;209 (3) life insurers faced informal limits,usually precluding ownership of more than 2% of an industrial firm; 210 and(4) during Britain's pre-Thatcher socialist era from the 1950's until 1979investment managers, when taking large and visible positions in industry, hadreason to fear nationalization.2 '

Thus, to understand Britain one must discount bankers and insurers due toinformal historical constraints (common in England and frustrating to research)and a socialist system (until 1979) that nationalized large enterprises. Thesepartial restrictions have eroded and British ownership of large firms ismoderately concentrated.212 British institutions, particularly insurers, exercisesome voice, although less than the current legal limits would allow, andcertainly less than their German and Japanese bank cousins.213 The Britishinsurers' greater voice and concentration indicate that restrictions and theirhistory matter; however, the fact that the British institutions have not adopteda German or Japanese structure suggests that their developmental path makesmassive structural change difficult, or at least that factors beyond economics,law, and politics, must be added to complete our understanding of corporatestructure. Law's historical effect on intermediaries may be less that of a damthat keeps intermediaries and firms apart than that of a blockage that divertsthe intermediaries into another path.

208. Tilly, supra note 141, at 196-98.209. Christine M. Cumming & Lawrence M. Sweet, Financial Structure of the G-1O Countries: How

Does the United States Compare?, FED. RESERVE BANK N.Y. Q. REV., Winter 1987-88, at 14, 15; LorettaJ. Mester, Banking and Commerce: A Dangerous Liaison?, FED. RESERVE BANK PHIL. BULL., May/June1992, at 17, 21.

210. LAWRENCE D. JONES, INVESTMENT POLICIES OF LIFE INSURANCE COMPANIES 103 n.74 (1968).211. Cf. HARGREAVES PARKINSON, OWNERSHIP OF INDUSTRY 1-2, 52-53, 102-104 (1951) (examining

relationship between ownership structure and risks of industrial nationalization); John Farrar & MarkRussell, The Impact ofInstitutional Investment on Company Law, 5 COMPANY LAW. 107, 109, 114 (1984)("[British i]nstituional investors are worried about the political consequences of an exercise of power. Theyeschew public criticism and fear public intervention.").

212. Franks & Mayer, supra note 95, at tbl. 2 (64% of the 200 largest British firms have a stockholderowning more than 5%).

213. Black, supra note 166, at 928 ("British... insurers, are quite active by American standards.");Getting rid of the boss, ECONOMIST, Feb. 6, 1993, at 13, 13 ('To some extent, British [CEO's] have alwayshad to answer to the tightly knit world of the City of London, whose old-boy networks and customs oftenserved as a check on their worst excesses. But the [CEO's] of big American firms were true masters of allthey surveyed.").

199319931


C. The Paradox of International Competition: Solution and Problem

Managers are situated in several markets and organizations. The internalorganization of the boardroom and its relationship with institutionalshareholders is only one. Professional pride of managers and directors makesthem try hard even if the organizational constraints acting on them are weak.Embarrassment from media attention will help correct the most egregiouserrors. Product markets, capital markets, managerial labor markets, employeelabor markets, and corporate takeover markets constrain managers. This is allanother way of saying that corporate governance is only one dimension ofcompetition.

Until recent decades, large-scale American industry had two criticalcomplementary advantages over much industry abroad: scale economies andcompetitive structure. The American market was (and is still today) so largethat it could support two or three firms, and hence workable competition, atthe highest economies of scale in even the heaviest of industries; no othermarket in the world could do both. Economies of scale in a small nation meantmonopoly; competition often meant inefficient scale. America could get bothscale and competition; so, if some details in organization at the top and in theboardroom were not for the best, no matter, because scale and competitionwere so important and might allow for profits that would hide a few defectsin organization.

These advantages are no longer exclusively America's. A common marketin Europe and a globalized marketplace allow foreign firms scale economiesin a competitive market. Details that were once trivial-like corporategovernance-become noticed in the 1990's. Stress from internationalcompetition reveals weaknesses in some American firms, presses Americanfirms to change, and makes us wonder whether international competition is notjust among firms, but also among forms of organization. z 4 The existence,persistence, and until recently, the success, of some rival foreign firms makeus question whether the American system is as sound in all dimensions as wehave thought.

Competition also partly substitutes for effective corporate governance.Successful foreign firms compete with American firms, pressing badlygoverned American firms to change or to lose market share. Response to thestimulus of product market competition generally induces American firms tobe more productive; corporate governance becomes secondary, relevant onlyto how (and if) it facilitates change inside the firm.

214. In this sense, international competition mirrors Tiebout's model of state competition as a meansof providing efficient regulation. See Charles M. Tiebout, A Pure Theory of Local Expenditures, 64 J. POL.ECON. 416, 419-20 (1956).

1994 [Vol. 102: 1927

Corporate Structure

There is another response that American firms might have to the stimulusof competition, other than to improve. In industries with long-lived capitalresources, poorly governed, poorly performing firms may stagger along, ashave steel and auto firms, wasting fixed assets without really improvingenough. Troubled firms may also respond with political action to reducecompetition-lobbying for quotas and tariffs-as have steel and auto firms.

CONCLUSION

The American-style large firm, with managers at a powerful pinnacle inthe hierarchy, is not inevitable. Managers could share authority withintermediaries holding large blocks of stock as they do in Germany and Japan.These foreign differences contradict parts of prevailing corporate theory, whichexplains key features of the firm as reactions to the principal-agent problemsthat arise when the firm, to reach economies of scale, sought large inputs ofcapital from thousands of far-flung investors demanding diversification. Asinvestors became more fragmented, power inside the firm shifted from theshareholders to the managers. While these capital-raising and principal-agentproblems are universal, the resulting corporate structures are not; institutionalarrangements other than those prevailing in America are possible, and do exist.

There is more than one way to move savings to industry-the Americanway, via weak intermediaries in a well-developed securities market is only one.Another is via strong intermediaries, like those in Germany and Japan, whichnot only move funds to firms, but also share power with managers. Thestrength of financial intermediaries is a product not just of economic evolution,but of history, politics, and culture. As I have argued before, American politicsinduced laws that fragmented institutions, their portfolios, and their ability tocooperate, precluding the institutional alternative. In Germany and Japan, theownership structures are different, the corporate structures in which CEO's areembedded are different, the financial institutions are different, and the rulesgoverning financial intermediaries are different. The foreign relationshipswould be illegal here. We will never know for sure whether American bankswould have developed differently without the restraints. We do know thatAmerican-style restraints would have been sufficient to stymie banks fromforeign-style involvement in corporate governance.

The classical model could be modified in two ways. One would be to scaleit back, by conceding that financial rules condition the way corporations grow,so that the American features are the best contractual arrangements availablegiven the restraints. The other way to modify the economic model would beto expand it by arguing that ownership in Germany and Japan will fragment,due to the inevitable economic imperatives of organizing financialintermediaries. Inevitable fragmentation may yet become a true prediction, butit is now only a prediction, not a fact. In Germany, concentrated institutional

199519931


ownership has been increasing, not decreasing, and new alliances that imitatekeiretsu relationships belie fragmentation. And Japanese concentratedownership has been rock-solid for twenty-five years, despite enormous pressureon the banks. Newly-rising, large Japanese firms have adopted the oldconcentrated ownership, not shunned it.

Political forces in Germany and Japan influence intermediaries andcorporate structures, although the forces have been different. Germany's statistpolicies concentrated banking at the end of the nineteenth century. Laterpressures produced codetermination, which is hardly apolitical and created avery different corporate structure than that which prevails in America. Alongwith German populism, it induces some banker disengagement today. Japaneseregulatory politics concentrated Japanese banking, separated banking fromsecurities, and then forced savers and firms to meet in the banking channel.Today, the Japanese system is under stress, but interest group infightingbetween bankers and brokers stymies full reconstruction of the financialsystem.

Firms in nations that historically tolerated or encouraged large pools ofprivate economic power evolve differently than do firms in nations thatrepeatedly fragment financial institutions, their portfolios, and their ability tonetwork blocks of stock. Thus, some of today's incipient fragmentation abroadcan be traced to political influences. Moreover, the current economic turmoilin both Germany and Japan makes us wonder whether there will be economiclosers associated with the large firm-employees or managers or both-andwhether they will appeal to their legislature to re-do the economic result inways that affect intermediaries and the structure of the large firm, replaying theAmerican political influences that created the American large firm.

Differences do not imply superiority. Labor, regulatory, or culturaldifferences could explain good performance with the differences in corporategovernance spurious. Although corporate differences suggest the importanceof asking whether institutional voice can improve firm governance, theevidence of foreign superiority is thus far scant. The inquiry here is by nomeans the final word; we have only begun to scratch the surface ofcomparative corporate governance. The direction for the future, I believe, is tounderstand more than just how different structures could make managers moreaccountable, but also to see whether institutional voice could addressinformational defects in securities markets and defects in organization,including industrial organization, labor organization, and the psychology oftop-level decisionmaking.

Even if we had more data, the foreign systems would yield no blueprint,because bankers, the protagonists abroad, are not likely to be the key activeinvestors here. Other financial institutions will have to be the source ofinstitutional voice in America, and we cannot get detailed blueprints for themfrom abroad. While there may be some value to removing the impediments to

1996 [Vol. 102: 1927

Corporate Structure

institutional organization, we have little reason to copy the foreign bankingsystems' details.

This analysis demands a broadening of the corporate academic agenda.Rather than using agency costs or contract theory or judicial doctrine toexplain this or that feature as mitigating or reflecting managerial deviationfrom the maximization of shareholders' wealth, we must consider the role ofpolitics, history, and culture. I lack the deep knowledge of German andJapanese law, corporate structure, history, and politics needed for a definitiveinterpretation. Although others will have to extend my analysis, the evidence

now available casts doubt on the economic model as a full explanation for theBerle-Means corporation. The key lesson is that the prevailing economic modelcannot completely explain corporate forms abroad.

Thus, I examine Germany and Japan not to argue that their corporatestructures are better and should be mimicked, but to show that different

structures are possible; that the American pattern of fragmented shareholders

with little power is not inevitable; that managers can share power with

intermediaries without making the corporate world fall apart; and that corporatestructure is highly sensitive to the organization of financial intermediaries,while the organization of intermediaries is highly sensitive to a nation'spolitics.

199719931


APPENDIX2 15

TABLE VIII. Percentage of Stock of Largest German CorporationsHeld Directly by German Banks216

Rank and Name Name of Bank Percent ofof Company Outstanding Stock

1. Daimler Benz

10. Thyssen AG

14. BMW

16. Deutsche Lufthansa AG

21. M.A.N. AG

23. Messerschmitt B61kow-Blohm

25. Karlstadt AG

35. Kaufhof AG (1986)

38. Degussa AG

40. Vereinigte Elektrizitdts Westfalen

42. Metallgesellsehaft AG

46. Preussag AG

57. Linde AG

58. Klockner Werke AG

67. Hochtief AG

71. Kldckner-Humboldt AG

75. Philipp Holzmann AG

90. PWA Papierwerke AG

92. Bergmann-Elektrizitiits AG

Deutsche Bank

Dresdner Bank

Commerzbank

D. Genossen-Bank

Bayerisehe L-Bank

Commerzbank

Dresdner Bank

Bayerische L-Bank

Commerzbank

Dresdner Bank

Bayerische V-Bank

Deutsche Bank

Commerzbank

Dresdner Bank

Commerzbank

Dresdner Bank

Deutsche Bank

Westdeutsche

Deutsche Bank

Dresdner Bank

Westdeutsche

Deutsche Bank

Commerzbank

Deutsche Bank

Commerzbank

Deutsche Bank

Deutsche Bank

Commerzbank

Bayerische H&W

Deutsche Bank

Bayerische V-Bank

28.20

1.60

1.600.30

1.60

4.94

5.00

5.00

4.94

5.00

5.00

>25.00

>25.00

9.00

3.00

10.00

6.30

>7.20

10.72

23.10

48.80

10.00

10.00

19.60

12.50

41.40

35.40

5.00

25.00

36.50

25.40

Source: DEUTSCHER BUNDESTAG, supra note 16, at 202-06.

215. Tables I-VII are in the text.216. Largest 100 Companies in 1988. Excludes stock that banks hold through mutual funds or as

custodian.

1998

Corporate Structure

TABLE IX. Relative Size of German, Japanese,and American Banking Sectors

Assets of German Banking SectorGerman GNP

Assets of Japanese Banking SectorJapanese GNP

Assets of American Banking SectorAmerican GNP

$1,900B$1,300B

¥728,577B¥436,927B

$3,399B$5,465B

- 146%

- 167%

= 62%

Source: MONTHLY REP. DEUTSCHE BUNDESBANK, Feb. 1991, at 4 (size ofGerman banking system); Memorandum from Bank of Japan to Mark Roe(Aug. 13, 1991) (on file with author) (size of Japanese banking system); FED.RESERVE BANK N.Y. BULL., June 1991, at A18 (size of American bankingsystem).

The German and Japanese banking sectors are about two or three timeslarger than the American, although the largest German and Japanese banks areapproximately five times larger than the largest American banks. The relativeweakness of the American banking sector accounts for roughly half of therelatively small size of American banks.

19991993]


TABLE X. Relative Size of Largest Bank and Largest Firm in Germany,Japan and America

Assets of largest German Bank - $267B =Assets of largest German Automaker $49B

Assets of largest Japanese Bank Y435B =

Assets of largest Japanese Automaker ¥55B

Assets of largest American Bank $216B 1.2Assets of largest American Automaker $180B

Source: Bank Data from Table V; Automaker Data from WORLDSCOPE (W/DPartners, ed., 1990), available in NEXIS, Compny library, Wldscp file(Toyota's assets are at June 30, 1990; all others are at December 31, 1990.).

This relationship shows the relative ability of a nation's largest bank tocontrol a large portion of the capitalization of the largest industrial company.We see that American banks are not as able to control large pieces of thelargest firms, and there would be concern that the American banks would bepoorly diversified if they did. The largest German and Japanese banks, incontrast, are able to control larger pieces of the largest industrial companieswithout committing as much (as a percentage) of their assets and equity aswould be required of an analogous American bank.

Corporate Structure

0)r-0 04 r- w t - 0 0 - 0

Ci In Ci .i .i . . . (. - Ci

n % 0 0% C) C- - 0 0 00 0% 0cl - - - Ccic 4- c

NI. C i c C i c C14 Ciq C14 i ci

"8, ca 12 2- % 0%

N - i "i cI- - - - - -" i

N

0 i - - - - - - .

C" 80 C C"4 000 000 0% l 0 4

0) 0% 0% 0 0% % 0% 0% 0 ~ 79

1993] 2001

00-

S0O0

V)<CO

°.. .co

,U

.)

'0

'0

C,.) ,10

'0 >,

So

<0..

PQ

0zo <


0D 6% 0 0 0)C 0

C) ) a 0 0) 0) 0

S00000000

cz

.0 N 02 M 20d 02d0

mC. m - 0 0 0 0

2 D 0 000

C' 6i 6 6 6 6 6 6U

00 to ' '00000

" O0000000

10

o 0c0 0 o 0 0

In CC C) 0 0 0) 0 0) 0 0

. 6 6 66 6 6

C2 0 0 02 0 0 02) 02

.00 0 -0 0 0 0 0 0 0%

CON

S00000000

0d0000000

0

ooo66oooo

0

0 G- CCo'

Q 'N cl

.3\

0

.o

o• O .,-.-

0 M-

o Ho .l

0U 0

r- 44-CZ

U . ... . ..,

0

>- 00C

0 '- oo .

CU ,. O\ H--

4- t C' o

0 M

- '.C~,.. -

zo. C. s ,

2002

Corporate Structure

TABLE XIII. Ownership (Voting) Structure for Three TypicalAmerican Large Public Firms (1990)

Source: Brancato et al., supra note 10.

Exxon IBM General Electric

Institutional Percent of Institutional Percent of Instutional Percent of

Manager Shares Manager Shares Manager Shares

CREF 1.09 Wells Fargo 1.07 Wells Fargo 1.09

Wells Fargo 1.05 Mich. St. Treas. 1.04 Bankers Trust NY 1.02

BankersTrust NY 1.01 Bankers Trust NY 1.01 CREF .99

NY St. Common Ret. .88 NY St. Common Ret. .83 Mellon Bank .91

Mellon Bank .73 Mellon Bank .76 NY St. Common Ret. .88



1993] 2003

Documents

Some Differences in Corporate Structure in Germany, Japan